John's Alu Profile build

[Evengravy]

Hi All,

So I'm starting this thread quite a way through the build process. I didn't want to document the entire thing from the start as I knew this was going to take me a long long time.

I built myself a Momus CNC build some years back and got the bug, but that machine is pretty limited as it flexes all over the place (on the Z axis). It's also belt driven which is an absolute pain in the backside as the belt pulleys loosen over time. I initially wanted a machine for wood and after time started machining more and more aluminium, so the new machine I want to be two fold: rigid enough for working aluminium but with a much bigger footprint for working on larger hard wood projects. I also do quite a bit of diamond drag engraving on aluminium which the old machine isn't accurate enough to do.

I basically have been piecing it together bit by bit over the last couple of years, and at the current stage it's moving and wired, but I have some changes to make for sure.

The Gantry is 80 x 160 profile and its running on two 80 x 120 profiles in the Y axis.

Linear rails are all 20mm THK ( the black Y axis rails are Raydent(?) coated). I got all the higher end motion parts used (but they're in great shape).

The Z axis was a real Achilles heel of my old machine, terrible design, so this new Z axis was bought as a full unit from an old machine in Korea (used for IC production I believe) it is made from stainless for the most part and has 15mm THK rails and a THK ground ball-screw for movement and optical limit switches. I milled an aluminium backplate for it and mounted it to the Y axis blocks.

I have a Ground NSK ball-screw on the X axis, positioned on the top of the extrusion. I chose to do that as I wanted to get the Z axis as close to the rails as I could.

The Y axis is driven both sides with ball-screws (Chinese 1605) but I'm really not happy with those so I just ordered some 20mm (2005) ground ballscrews from TBI. I would have liked something better brand wise but I think they should be okay.

Most of this is probably overkill in a aluminium based router I know, but I want to get it as functional as I can really.

You'll notice I haven't got a base/table as yet, the reason being is I'm a bit undecided about that. I've been toying with the idea of mounting it to a wall (and have some blocks ready for doing that, but I'm struggling to decide. I really don't yet know how I could effectively level the wall surface in practice before fixing the machine to it (any advice would be greatly appreciated.)

The spindle looks like the regular cheapo chinese ones but I went with a company Jain Ken for this, it's 1.5Kw ER16 and has ceramic bearings, as far as I can measure it has no runout on it (!). Mind you my dial indicators aren't serious quality, but it's definitely accurate enough for my needs.

For VFD I went with RS510 from RS electronics, it works really well and is essentially a Teco 510 it seems.

I made up the control box from scratch, a linear power supply (70V) was made with a transformer, regulator plus capacitors and I put a soft-start on the incoming AC to smooth out the turn on a little.

The stepper drivers are all AM882 running at 8th micro stepping. I'm really impressed with these drivers, they're very smooth.

I'll figure out how to upload some images shortly, and add them here.

[Evengravy]

Here's a picture of the main machine as it is today.
27712

[Evengravy]

Here's a couple of shots of the spindle and Z axis details.
277132771427715

[Evengravy]

The Z axis came with three optical sensors inside, I kept one in there as upper homing limit switch and reused the other two as home/limits for the Y axis, one each side. For mounting I machined a carrier from Delrin/Acetal. I'll probably replace this with aluminium at some point soon.

27716

The plates which mount between the Y axis bearing blocks and the gantry were modified to fit a piece of aluminium L profile, this moves into the optical switch when homing.

27718

For mounting the Gantry to the ballscrews on the Y, I had a local company weld me up some brackets to my design.

27719

They made a bit of a balls of it to be honest, It's not square and I had to rework it by hand a lot. I can't weld so I didn't have much option, I really should learn to. Since I'm replacing these chinese ball-screws with ground screws from TBI I'm not going to sweat it just yet. Once the parts are here I can redesign this part and the motor coupling for each.

I'm using LinuxCNC 2.8 for this with kins, if anyone wants me to upload my .ini and .hal files let me know. It's working well so far.

[Evengravy]

So, as I said the spindle is from China, but seems to be much better quality than the regular ones, certainly run out and machining quality is amazing. Running at full pelt it is really smooth.

Here's the company if anyone is interested (no affiliation with them at all btw): http://www.jian-ken.com

The VFD is model RS510 from RS Electronics (no affiliation with these either): I'm really happy with the quality of this too. It's actually rated for 2.2Kw but I thought some overhead would be a good idea. Definitely one to add to the list for the other Chinese spindles too.

27720

[Evengravy]

Part of the reason it's taken me so long, apart from sourcing all the parts here and there, is the fact I had to build myself a little workshop to house it in.

27721

I still have a bit to do to finish this place, insulate it in between the stud walls and plywood the internal walls, but it's dry and solid.

At the right hand side outside you can see the six inch blocks. I don't know if these exist in the UK or not but my idea was to build a wall with these on the flat and bolt the entire CNC to it with the Y rails running parallel to the floor. If anyone has an experience I'd appreciate some advice.

27722

I'm really not sure how I would level a wall to the degree necessary to accurately fit this too, but maybe I'm overthinking it.

[Evengravy]

As a test, I knocked up a pencil holder out of some delrin and some drill rod, with a spring for applying pressure and for accounting from deviations in the surface. It worked great, but the chinese ball-screws really aren't great (noise, vibrations etc.)

27723

I actually sourced angular contact bearings for the BK12(?) blocks, so I might now have to sell those too.

[Evengravy]

Here's a snap of the internals of the control box, the main cabinet was sourced from RS electronics and I fitted the rest after bench testing. I probably could have used a smaller case but I have plenty of space so I don't mind the extra space from a cooling perspective.

27727

I went with GX 16 type pins for termination of the wiring both on the motor/sensor side and the case/enclosure side. The wiring is done so that the case side takes the Sheild of each wire to earth.

27728

I think that's all I have for now guys/gals.

[Evengravy]

A couple of questions I have, then, if you don't mind:

+ Playing with the machine as it stands, there's a lot of resonance from the hollow sections of the profile, for the Y profiles I intend to either fill with sand or maybe epoxy/granite. With one end almost closed at the front and the significant vibrations from the less than stellar 1605 ball screws these are resonating like an instrument at the moment. I'm sure this will hurt surface finish as I can clearly feel the transmission through to the spindle by hand, but I expect a sand fill will help since these parts won't need to move.

+ With regards the Gantry, I'd imagine filling that with epoxy/granite will be much too heavy for my motor/screw combo (480oz/in plus double nut 2005 ground ball-screws). Has anyone any experience with filling the voids of profile with expanding foam to dampen vibrations/resonances? I suppose I could test a piece but I don't have any spare material to hand. I wonder if something like silicone fill could work if it could be done cheaply enough.

+ Also would be keen to hear any input from the idea of bolting this entire CNC to a concrete wall, specifically how I might line everything up in practice. I like the idea of gravity pulling away chips and using less floor space but could just go for the traditional setup.

[mekanik]

Hi John
Regarding vertical mounting you will have to build yourself a substantial base, if you make another post asking for ideas for designing a vertical assembly machine Dean(JAZZCNC) might give you some advice regarding this, he has a video of one of his larger builds but i couldn't find it.
Regards
Mike

[Evengravy]

Hi John
Regarding vertical mounting you will have to build yourself a substantial base, if you make another post asking for ideas for designing a vertical assembly machine Dean(JAZZCNC) might give you some advice regarding this, he has a video of one of his larger builds but i couldn't find it.
Regards
Mike

Okay, thanks a lot, I'll have a think about some options.

[Kitwn]

John,
I haven't tried this but have wondered if polystyrene concrete could be used to damp the gantry without it getting too heavy. I found an article once (don't have it to hand I'm sorry) that described using up to 60% by volume of polystyrene beads (you can buy them for making bean-bag furniture) to make concrete for light-weight structures. It's still quite strong at that percentage and I think the density came out at a little over 1Kg/litre.

[Evengravy]

John,
I haven't tried this but have wondered if polystyrene concrete could be used to damp the gantry without it getting too heavy. I found an article once (don't have it to hand I'm sorry) that described using up to 60% by volume of polystyrene beads (you can buy them for making bean-bag furniture) to make concrete for light-weight structures. It's still quite strong at that percentage and I think the density came out at a little over 1Kg/litre.

Hey. Thanks! That's quite interesting. I was thinking of something similar with perlite and concrete or epoxy. I was told that concrete may react with aluminium though somewhere but I can't recall where exactly. Does a anyone have any insight on that? On this note I was looking at some local companies that sell recycled tyres, chipped up for playgrounds. I was thinking that could make a good aggregate for a damping composite maybe, with either epoxy or liquid platinum cure silicone to bind the chunks together in a similar way to making epoxy/granite.

It sound like some tests might be in order, luckily I'm an acoustics specialist with oscilloscope etc here so could at least test a few solutions on a small scale.

[JAZZCNC]

Hi John
Regarding vertical mounting you will have to build yourself a substantial base, if you make another post asking for ideas for designing a vertical assembly machine Dean(JAZZCNC) might give you some advice regarding this, he has a video of one of his larger builds but i couldn't find it.
Regards
Mike

That would probably be this 8x4 then. 10 x 5 with a Vacuum bed will be coming soon if I ever get out of this bloody house again.!!.

The vertical machine works amazingly for large machines with no downsides other than cannot use it as a bench. The advantages are many and I wouldn't go back to a horizontal machine if larger than 4x4. That said I'd still have a vertical machine under 4x4 for some of the hidden benefits like longer tool life and better chip management but below 4x4 the space-saving isn't so great.

Regards the filling of tubes then I've tried sand and while it does help a little it's not a massive difference. Don't fill tubes with concrete because it can shrink, it also doesn't mix well with aluminum. Expanding Foam and Silicone etc don't do diddly squat so don't waste your time money.

The best thing I found is to use a thicker wall tube in the first place and brace it well. Regards the aluminum profile then if you use HD version then you shouldn't have any issues with vibrations and I don't think to fill the voids with Epoxy granite will dampen to any great degree but it will certainly affect performance because of the extra weight.


https://www.youtube.com/watch?v=wJB_75bLEIM&t=8s

[Kitwn]

It sound like some tests might be in order, luckily I'm an acoustics specialist with oscilloscope etc here so could at least test a few solutions on a small scale.

You're the man then! I like the idea of bendy concrete made with old tyres :excitement:

[Evengravy]

That would probably be this 8x4 then. 10 x 5 with a Vacuum bed will be coming soon if I ever get out of this bloody house again.!!.

The vertical machine works amazingly for large machines with no downsides other than cannot use it as a bench. The advantages are many and I wouldn't go back to a horizontal machine if larger than 4x4. That said I'd still have a vertical machine under 4x4 for some of the hidden benefits like longer tool life and better chip management but below 4x4 the space-saving isn't so great.

Regards the filling of tubes then I've tried sand and while it does help a little it's not a massive difference. Don't fill tubes with concrete because it can shrink, it also doesn't mix well with aluminum. Expanding Foam and Silicone etc don't do diddly squat so don't waste your time money.

The best thing I found is to use a thicker wall tube in the first place and brace it well. Regards the aluminum profile then if you use HD version then you shouldn't have any issues with vibrations and I don't think to fill the voids with Epoxy granite will dampen to any great degree but it will certainly affect performance because of the extra weight.


https://www.youtube.com/watch?v=wJB_75bLEIM&t=8s

That's awesome! This has given me some ideas for sure... Design wise it's not a million miles away from mine, I really wish I could weld. Maybe I can think of something I could make work using aluminium profile.

Ten four on the cement filling. I think I'll go with sand on the Y rails that don't need to move just for the added benefit, and try some test pieces for composites on the X gantry, to see if it's worth the hassle. I can't recall if it was HD rails, it was this type of profile from Dold Mechatronic: https://www.dold-mechatronik.de/Aluminum-profile-80x160L-I-type-groove-8 maybe there is heavier type out there and that was my mistake.

[Evengravy]

You're the man then! I like the idea of bendy concrete made with old tyres :excitement:

Thanks. Sure, I'll do some more in depth reading and maybe some tests soon, the complication at the moment is getting out of the F'ing house to get materials but gives me chance to do some research if nothing else.

I came across a guy on another CNC forum that recommended bitchumen (tar) plus sand, that could potentially be a cheap and available solution, if it isn't too heavy that is. Maybe roofing tar plus rubber chips. Certainly it's common to use roofing bitchumen sheets (old school roofing felt) in acoustic dampening solutions in a studio environment, I see no reason why it wouldn't transfer here, in theory.

Silicone would likely be expensive but should be quite efficient, maybe recycled rubber plus a binder might help. It's always going to be a tradeoff of weight vs dampening factor for a gantry though.

It's simple enough to test smaller pieces in practice though; with a contact mic/measurement mic and some shorter pieces of profile and an oscilloscope, I have some 80x80mm profile I could donate to the cause so I could side by side a few possible solutions as an experiment. Could be useful to see what approaches might work down the line.

At the moment, I'm just awaiting the new ballscrews to arrive so I can get rid of these cheaper ones.

Thanks for chipping in guys, appreciated.

[JAZZCNC]

That's awesome! This has given me some ideas for sure... Design wise it's not a million miles away from mine, I really wish I could weld. Maybe I can think of something I could make work using aluminium profile.

Welding isn't that hard, esp if you can get your hands on a Mig welder. You can even hire them from some a decent hire shops.
The welds don't need to be industrial strength. Steel is very forgiving so it allows you a few attempts if get it wrong and along with a grinder you can hide all your screw ups away.


I can't recall if it was HD rails, it was this type of profile from Dold Mechatronic: https://www.dold-mechatronik.de/Aluminum-profile-80x160L-I-type-groove-8 maybe there is heavier type out there and that was my mistake.

That is the ITEM "L" section the "L" stands for light. It's exactly what you see on that vertical machine but I'm using the heavier version which doesn't end with "L".
However, I've used the "L" on many machines and it works perfectly fine so I wouldn't be too worried.

[JAZZCNC]

If you are going to do vibration tests then I wouldn't waste time testing individual pieces of profile or steel etc off the machine because all that matters is the tool. Testing at the tool is the only thing that matters IMO. You'd be better investing time on designing a strong Z axis.

[Evengravy]

If you are going to do vibration tests then I wouldn't waste time testing individual pieces of profile or steel etc off the machine because all that matters is the tool. Testing at the tool is the only thing that matters IMO. You'd be better investing time on designing a strong Z axis.

Yea, that's a good point. I did a bit more reading around the subject and a guy on CNC zone recommended pea gravel in the non moving parts and nothing in the gantry, for now. I'll try that first and see where it is. I actually have a MIG that I bought years ago but never got around to learning to use the thing, a gas-less 150 from draper (I'm not sure if they're much good). It's been shelved in my dads place for years but maybe it's time to dust it off.

I think the Z axis I have is pretty solid though, it's foundation is milled from stainless. I know the opposite arrangement than most recommend, with the rails on the z plate which seems to be preferred, but it was a simple solution for me. I've tried to keep everything as close to moving bearings and having as little overhang as possible. The Z only has 100mm of travel but after using my other machine for a few years I realised I've never needed to mill anything over a couple of inches thick so the z overhang was really compromising the strength. That plus the momus design is not very good generally. Live and learn. I thought it would be a good idea this time for me to keep the gantry and Z axis as tight to the part as is feasible.

It's a shame about the profile, maybe I'll change it for something heavier but at this stage I might just run with it and see how it does, should be easy enough to swap it out for a heavier profile down the line.

[JAZZCNC]

I think the Z axis I have is pretty solid though, it's foundation is milled from stainless. I know the opposite arrangement than most recommend, with the rails on the z plate which seems to be preferred, but it was a simple solution for me. I've tried to keep everything as close to moving bearings and having as little overhang as possible. The Z only has 100mm of travel but after using my other machine for a few years I realised I've never needed to mill anything over a couple of inches thick so the z overhang was really compromising the strength.

There's nothing wrong with rails on the rear plate, I've built dozens of machines using both methods and neither is better than the other because the difference between them is hardly anything and any advantage one as over the other is mostly dependant on machine design and user needs.
In your case with a short Z-axis, it's the best method because it allows you to make the most out of a short-stroke while allowing you to keep the gantry low and squat giving the best strength.

Edit:

Don't worry about the profile it will be fine. If you using my "L" Gantry design just make sure you bolt it together well.

[Evengravy]

There's nothing wrong with rails on the rear plate, I've built dozens of machines using both methods and neither is better than the other because the difference between them is hardly anything and any advantage one as over the other is mostly dependant on machine design and user needs.
In your case with a short Z-axis, it's the best method because it allows you to make the most out of a short-stroke while allowing you to keep the gantry low and squat giving the best strength.

Don't worry about the profile it will be fine.

Ah that's good to hear, I didn't go for the L design you have no, I wish I had of kicked around here before but that does look like a better solution than mine, unfortunately I'm quite a way through the build so at this point I'm just looking to get her done and running. I don't know if you would have any comments on the design as it stands? There's a few places I could reinforce around the gantry connection to the Y axis plates etc. I'm sure I can figure something out there.

I could always beef up the gantry with some steel plate on the back-side too, if that would help. I might just get the base/table done next and see how she does though.

[JAZZCNC]

I could always beef up the gantry with some steel plate on the back-side too, if that would help. I might just get the base/table done next and see how she does though.

I would look to beef up and brace the gantry ends but other than that it's nice and stiff looking.

That Z-axis is spot on for the size and you have half-decent material size. You won't get a much stiffer setup.

NOW just turn it on its side, or stand on its end and bolt to that wall.!!! . . . . You'll love it.

[Evengravy]

Thanks mate I'll do that,

Once the steel shops open again I'll have a crack with the MIG and see how I get on.

[Evengravy]

So,

New ballscrews have arrived, TBI motion ground 2005 with double nuts. These ballscrews have BK15 mounts at both ends rather than a floating end and each BK15 holder, four in total, has a pair of Japanese 7002 angular contact bearings. These parts were bought lightly used from the same seller that I bought a lot of the other parts from in South Korea, but are clearly much better than the Chinese ones that I'm going to pull off the Y axis. One of these screws will be on each side of the Y axis as before so it should improve the smoothness of motion and the extra thickness of the screw (going from 16mm to 20mm) should enable me to increase the speed of the rapid movements a little without encountering whipping.

27781

The only downside of the change so far is that I've lost a bit of travel on the Y axis, down from just over 1100mm to just over 1000mm. I'd rather have the piece of mind of these motion parts than the extra cutting envelope so I'm okay with it.

As a result I've needed to redesign a few parts though. As per Jazz suggestion I've added a new bracing for the gantry which will tie in to the Y axis linear bearing plates. This part only braces towards the back side of the gantry, as the new ball-screw connection will be made to the front side via a plate as shown below. I might figure out something for the front too at a later date.

27779

I've designed some new shoes for the ballscrews to fix to to mate them with the Y axis profiles.

27778

I've also rethought the design for connecting the ballscrew to the Gantry. I was never too happy with the previous welded solution, they were a mess.

27777

Lastly, since I've moved up to a larger screw my motor mounts have to change too. These attach to the end of the Y profile on each side and carry the Y motors.

27780

Since I can't buy materials at the minute I've decided to try to get all of these laser cut in the UK and shipped over. The company are running but with minimal staff so it might take a while to organise.

Gantry to ballscrew mating part will be made of 15mm mild steel laser cut and I'll paint that later. The rest will be laser cut from 10mm stainless steel. That's the plan at the moment, it's hard to get companies to do much at the moment so we'll have to see.

[JAZZCNC]

Don't understand why you are making the Y ball screw plate and the brace plate separate.? This could be made as one piece and will be stiffer and easier to fit.!

[Evengravy]

Don't understand why you are making the Y ball screw plate and the brace plate separate.? This could be made as one piece and will be stiffer and easier to fit.!

Hey,

I designed it this way just to give me scope for adjustments in each degree of freedom. I couldn't think of a better way of doing it to be honest, open to suggestions though.

I could get both parts made up from mild/bright steel and once everything is lined up get the joints welded. I picked up my MIG today and dusted it off, maybe I can get good enough with some practice to get it done myself.

27785

Also there's scope to brace between the front and side section with some triangulated pieces.

27786

Honestly not too clear how I could make it in one piece, but I'm probably being blonde...it's been known to happen.

[JAZZCNC]

Hey,

I designed it this way just to give me scope for adjustments in each degree of freedom. I couldn't think of a better way of doing it to be honest, open to suggestions though.

Ah makes a bit more sense now. I was working on your old setup and seeing all those holes which matched the profile ends I thought was fastening into ends.

Carry on.!! . . Lol

Edit: The drop bracket still needs some bracing thou.!! or a redesign of the whole thing.!.

[Evengravy]

Edit: The drop bracket still needs some bracing thou.!! or a redesign of the whole thing.!.

Yea, I was planning some welded triangulation braces between the two parts once it's all aligned properly, the overhanging section that mates with the ballscrew is annoying me but I'll come up with something to strengthen it up. Front plate will be heavy (15mm) but I know that will be a weak spot for flex. I'll play around with the design a bit more in CAD.

[Evengravy]

Edit: The drop bracket still needs some bracing thou.!! or a redesign of the whole thing.!.

Hey, so I've decided to go back to my existing design of the gantry to Y ballscrew connection which was 10mm steel plate welded and braced with 10mm triangular plate sections. After testing this arrangement on the Chinese screws for play, measured at the tool with my dial indicators I'm quite happy with it strength wise so I have redesigned the same arrangement to fit the new ball-screws. The current layout has between 0.02mm and 0.03mm of play which is largely coming from the ballscrew nuts themselves, you can feel a good bit of play by hand in the ball nuts when these are off the machine which does tighten up significantly when both screws are in the machine (weirdly), but I definitely got unlucky with these Chinese ball-screws compared to what others seem to get. There's noticeable bends in them too so out they go. Bare in mind the current ballscrews are running with angular contact bearings in the BK12 holders, which I bought after the fact, so those bearings tighten up the play in the Y. Initially play was in the order of 0.1mm+ in my measurements before and fitting angular contact bearings. I can't feel any play at all by hand in the new ground ball-screws so I'm confident they'll be a big improvement.

One catch with the new design is that I'll need to machine a ball-screw nut mount for the new ball-screw, one each side. Would anyone here be able to machine this part for me (x2) out of Aluminium for payment? I'd rather spread work here than a bureau service if anyone is free/able/willing.

I don't think my Momus will be accurate enough to do it to this depth to be honest, there's just too much play in the Z (part is roughly 57mm x 65mm x 70mm from Ali (6061/6082 etc))

Failing that I'll have to consider milling it in thinner 25mm sections bolted together until I can machine it myself from a single piece when the machine is in situ, but I'd rather get it done right the first time if possible.

I still need to consider the bracing of the gantry to the Y bearing block plates, but I'm happy enough I can do that after the fact pretty simply and am considering an adjustable design there.

I'm leaning towards setting this up in the standard horizontal arrangement atm, reason being I'm thinking of running high flow flood coolant on it and it's easier to manage that way.

[Evengravy]

After getting no response from the laser cutting service on quotes I decided to have a route around in my stock and found some appropriate stock to mill some of the new parts myself.

Yesterday I went ahead and milled two new motor brackets with spacing to suit the new Y ballscrews. Material is 6082 Aluminium which machines rather nicely with a single flute carbide end mill, I just have to slow things down a lot on my Momus CNC and use small diameter cutters as the Z axis is prone to flex.

I can still achieve pretty accurate results though, it just takes time. I can't wait to have a stronger machine that can make use of adaptive tool-paths.

27810278112780827809

I hand tapped threads with the aid of my tapping guide, I've had some bad luck in the past with using the battery drill for tapping and since it's so difficult to find materials at the moment I was happy to take my time to avoid any broken taps/ruined parts.

Whilst I was at it I milled four new feet, which connect the Ball-screws to the profile. Stock is 10mm thick and is set up so I can align the side of the BK15 holder to the top of the profile. These are tapped with M6 thread in four locations for the BK15 to bolt to with 45mm socket head bolts. There are then two 5mm fixing points for fixing to the profile. I may hide some countersink bolts under the BK15's too, for now it's fine.

I went ahead and ordered two ball-nut carriers from a custom CNC bureau service, so I can keep moving on this. They should be here in a week or so I think.

[Evengravy]

So, Friday I managed to get some new laser cut steel through the post from a crowd in the UK.

These parts will replace the current steel brackets which connect gantry on both sides to the ball nut holders. I used this as an opportunity to try welding for the first time. Please don't hammer me on the welding, I know it's shocking but this was my fist go at it and I've learnt a few things. One my 100A mig isn't nearly powerful enough for material this thick and, two; buying cheap gasless flux cored wire is a big mistake. If anyone is interested the Super 6 flux cored mig wire is something to avoid for a beginner at least. It creates plumes of black smoke, spatters like crazy (yep polarity is correct) and doesn't pool up very-well. If were to run the MIG again, I'd probably look to get some lincoln gasless flux cored wire, it seems to be well reviewed. Seems a bit pointless for this project, in future I'll be sticking to trying to learn a bit of ARC welding as I'm mostly interesting in sticking thick pieces metal together.

Essentially I've built two of these and each is just tacked at this point and I intend to grind this mess back and get up to speed with the arc and come back to it. It does seem to be fairly strong though. Importantly it is square and key surfaces are flat so it a definite improvement on the last version a local company made, which were neither flat nor square (but welded better for sure). I'll be able to move along with the other bits when they arrive.


Custom ballscrew nut mounts are on the way, then we can mount and test these new ballscrews!

One I can confirm that everything lines up the way it should, I can use the steel from the old ballscrew mounts to practice some arc welding before coming back to these. At the moment getting steel is tricky. I intentionally didn't weld on the inside so I can tackle that later, after some practice.

279672796827969

[AndyUK]

I'd be very happy with that as a first welding attempt - nice one! But yeah MIG isn't going to be best for that material size - I'd be surprised if it has enough penetration.

What's the phrase? Grinder and paint make me the welder I ain't.

[Evengravy]

Thanks Andy! it seems to be strong enough but I'd guess it's penetrated around 1-2mm into the base metal, or so (looking at the front corners). I can pass over the inside with the ARC once I get the hang of it too so I'm not too worried at the minute, I was more nervous to stuff up the alignment to be honest but it's bang on where it needs to be.

I've been playing around with the ARC a moment ago on some scrap. I like it much more than MIG as a process so far, at least I can get a good pool going with it. Enjoying that a lot.

I'm hoping to weld up a steel frame of some sort so it's worth learning something new whilst I'm about the house a lot.

[Evengravy]

This morning two custom ballscrew nut mounts arrived from the CNC service, they came out really well so I followed on to test fit these and the newly welded (ahem) brackets. 28020

With a little to and fro to swap parts out I'm glad to say it all fits as well as I can expect, and boy does the thing feel much more solid now.



I had to massage the brackets a little on the corner to give some wiggle room by adding chamfer the corner of the L to give a millimeter or two clearance.

28021

I'm now just awaiting the new anti-backlash couplers to mate the motors to the new screws and this part should be done. It's been a bit of extra work, but so far I can feel the improvement and I'm glad that I took it on.

If everything goes smoothly then I'll be removing these steel brackets and adding some more internal welds just to beef them up a little before painting them black and refitting. I'll need to start designing the table soon, I may be back to get some advise from you all when I get around to that part.

28022 28023 28026

[Mister Mills]

Hi John, It's looks great. In relation to the resonance you are experiencing in the Y axis are you running the machine with the Y axis unmounted to a surface or base? Do you only experience this when running the machine with the spindle on or off? Your selection of steps per revolution will also come into play with resonance as I myself experienced something similar. Endless tweaking and trying different combinations of steps per rev and mucking about with acceleration and deceleration settings got rid of it for me. Where are you based in Ireland? Liam.

[JAZZCNC]

I'd be very happy with that as a first welding attempt - nice one! But yeah MIG isn't going to be best for that material size - I'd be surprised if it has enough penetration.

What's the phrase? Grinder and paint make me the welder I ain't.

Mig is more than capable of welding that material, but 100A isn't. It would need 250-300A and thick wire. The trick with Mig is setting the wire feed correctly.

Also when your butting 2 pieces like you have if you grind a chamfer on each edge where you want the welds it will give you a stronger weld as you have larger weld area, it also gives you a flatter weld with hardly any grinding required.

[Evengravy]

Mig is more than capable of welding that material, but 100A isn't. It would need 250-300A and thick wire. The trick with Mig is setting the wire feed correctly.

Also when your butting 2 pieces like you have if you grind a chamfer on each edge where you want the welds it will give you a stronger weld as you have larger weld area, it also gives you a flatter weld with hardly any grinding required.

Hey, yes sorry I was referring to my MIG in particular, I'm sure with more current it would work absolutely fine. Ten four on the chamfering, I'll keep that in mind for future.

[Evengravy]

Hi John, It's looks great. In relation to the resonance you are experiencing in the Y axis are you running the machine with the Y axis unmounted to a surface or base? Do you only experience this when running the machine with the spindle on or off? Your selection of steps per revolution will also come into play with resonance as I myself experienced something similar. Endless tweaking and trying different combinations of steps per rev and mucking about with acceleration and deceleration settings got rid of it for me. Where are you based in Ireland? Liam.

Hey, thanks. Currently no base, I'm still considering the best approach for building the base at the moment. Some are quite unconventional lets say but I want to run high flow flood coolant for milling aluminium so I'm going to think about liquid capture etc. I'm going to try and CAD design some ideas over the weekend. I'm a bit limited with the current lockdown in sourcing what I would need anyway so no stress just yet. The vibration was really just coming from the screws themselves without the spindle running, which weren't great quality and not straight either. So, partly crappy ball nuts with lots of friction and play, partly non straight ballscrews plus a coupled hollow cavity = resonance. I'm not too concerned with the resonance part, that's in my field of expertise (I'm an acoustics/audio specialist) more concerned with the smoother motion that the new screws will provide plus hopefully less play (I'll need to measure that again today). I intend to fill the Y rails with either sand, pea gravel or epoxy granite to get rid of that resonance. I'll probably go cheaper options first and see where it is. I can measure the natural resonant frequency pretty easily and try to tune around it too. I'll keep those points in mind for sure though on tweaking, thanks for that.

I'm based in Donegal buddy, you?

[Mister Mills]

I'm down in Co Meath for my sins.... Yes the Lock Down is a bummer but I believe we will be let out gradually from the 18th... Plenty of time to do design work! I forgot to ask what the cutting area of your setup is? It's a decent build for sure. I would 100% recommend flood coolant doing any serious metal machining or mist for lighter work otherwise you will get a pain in your ass replacing cutters! I settled on rack and Pinion after trying belt and lead screw. Both turned out to be disasters.No way I could afford 2.8 Meter Ball screws!! My Rig has been sitting idle for 2 years and only recently am I getting the Itch to start playing again.. The only bonus of the lock down!. Feel free to give me a shout it your stuck with anything. I have made shit load of mistakes building my setup but you live and learn and it may help you from falling into some of those pit falls!!!

[Evengravy]

Ah. Thanks a lot for that. Aye coming from a belt driven build too so something more stiff is in order this time. Yep the rack and pinion sounds like a perfect solution for that size (that's a big machine, jealous) this one has a cutting area of around 1100mm x 750mm ish but should be big enough for my needs. I do a lot of aluminium work and bits of hardwood so don't need anything too large. You never know i might go for an 8x4 machine down the line.

[Mister Mills]

Bigger machine needs a lot more space to be usable. This is my downfall. I may well downsize her and go for a 4 or 5 axis design. I'm getting the Itch !!!!!!!!!!!!!!.

[Evengravy]

Bigger machine needs a lot more space to be usable. This is my downfall. I may well downsize her and go for a 4 or 5 axis design. I'm getting the Itch !!!!!!!!!!!!!!.

I've had a smaller machine than mine for quite a few years, with a cutting footprint of roughly 500 x 380mm and have not found that to be too limiting generally speaking, so didn't want to go too large. If I were to go 8x4 I think I'd go vertical like Jazz's to save space and help with chip clearing. I want to leave some room in my workshop for some other bits too. I acquired an old CNC lathe (EMCO) that I'm upgrading and I'd like to build a small plasma table at some point (maybe 2x4). We'll get this one done firstly anyway.

A rotary axis would be great, if you could pull off a 5 axis design I'd be very interested to see that come together.

[Mister Mills]

I may well downsize to 1m x 1 m or 1m x 1.2m. Seriously Don't have the room and no real need for the full size setup. It took me 6 months to build the Big girl! and if I am to be honest I don't know why I went so big. The joy for me was building her but she has never had a full sized sheet of anything on the table!!!!!!!. If I go 1 x 1.2(Cutting) I will arrange it so I can slide a full size sheet into and through her to save cutting it into several smaller pieces. I have a few smaller ball screws that might fit the new proposed dimensions and I would prefer to use those instead of the R&P. This time round I will Catalog the build :). I have some holiday time booked in 2 weeks so I might start the strip down then. Fingers crossed!

[Evengravy]

Ah here. It never ends. But I bet the next one you build will be better. My first one was an mdf abomination, second much better, and this much much better (so far) we'll see what happens. I'd never go away from propper linear rails again though, anything else is just a false economy in my book as you spend so much time messing you'd be better to buy something made for the job upfront and save the hassle.

I've been busy planning the base, but work is busy atm so I won't update just yet. Bits and pieces are on route but ill post info soon to that end.

[Mister Mills]

I totally agree about the rails. Mine came from a salvaged Bisse industrial cnc router. Top quality parts. Budget usually decides @ 3 meters long! I'm going to try and sell my 50W 600x500 laser cutter/engraver so that might help with the proper rail requirement. Doing a check list at the moment and I have everything I need. Says me as I blindly walk into this project. No doubt there will be something missing when I get to that point of" Shit I forgot about that! .
The size will be dictated by the ball screws i have (they came from a Salvaged Datron cnc machine) top notch. The 5 axis setup is really something I would like to do but I will start with 3 and keep the other 2 in mind when doing the new re-design. Good luck with the new base and be sure to post some photos.

[Evengravy]

Planning is in the works for the base, I've been collecting materials here and there as I can source them in the current climate. I've decided against the wall mounted option, when I drew it up in CAD it wasn't really saving me a massive amount of floor space (due to the relatively small footprint of my machine) and I really want to make use of high flow flood coolant so felt a regular horizontal setup would be best.

I've decided to be a bit unconventional with this design as outlined in the images.

2814428145

The main sides of the structure will be made from concrete based blocks, each measures 140mm(H) x 212mm(W) x 442 (L) and are very strong. These blocks are common here in Ireland and are largely made from a high proportion of granite aggregate along with a minimal amount of concrete as a binder (I'd estimate 15 - 20% concrete/cement) and I hope in this case, given the ratio of granite to cement, that shrinkage will be quite low and vibrations can be minimised too. I've had these blocks for over 12 months now so they should be relatively stable, but time will tell I suppose.

The intention for the main table surface is to cast this from home made glass fibre reinforced ultra high performance concrete. The intention is to cast this as a 75-100mm thick section in one pass by supporting the open section with ply braced from underneath (pre cut in the middle and siliconed at the joint for easy removal). I have sourced and received the glass fibre strands and the superplasticiser which I'll use to minimise the water content required in the concrete mix. I also intend to embed typical steel reinforcements (rebar or mesh or a combination of) I have yet to source those aspects as local steel outlets are closed for the next week or two. Time wise it's likely to be around the second week of June before construction can start, I'm looking forward to it.

For wiring/routing I intend to cast some 30-40mm plastic conduit (sink drainage pipe) into the front section, to allow motor and limit switch wiring to be routed cleanly from the left side to right. Controls box for the steppers and spindle VFD will mount to the exterior of the right hand wall.

I've been experimenting with some cast concrete of late and some alterations in mixes and proportions, likely I'll keep it fairly simple but I'll make some test samples soon to test those out.

The top of the table will be certainly quite rough post concrete casting, so the intention I have is to allow the concrete some time to settle and after that period level the entire table top with a 3mm layer of epoxy to give a nice flat surface to measure from and secondly, to provide some waterproofing. Part of the idea of designing the base this way is to allow sufficient space for a flood coolant capture tank to be housed in the underside of the 'table' and further design some fluid direction system once a tank has been sourced.

I could obviously go the tried and tested route and build the base from steel or aluminium, but I'd like to try this out to see how it goes. It's relatively cheap and if it all goes tits up, well I'm sure I can get a sledge hammer handy enough.

[Mister Mills]

Good stuff John. I would have never thought of a concrete base but now that I see you mock up it makes perfect sense. It will also be much easier to surround the upper part of the walls with a side/back etc to catch coolant. Flood coolant it a really good thing to have on any machine cutting aluminum. It will save toy a ton of cutters! Good luck with the build.

[Kitwn]

That's a really interesting base design, I've been wondering about a concrete or block base myself in the future as a cost-effective and very rigid structure. I'll be interested to follow your progress.

If you plan to drain the coolant to a tank underneath then it may help to cast a short length or two of plastic drainpipe through the base to act as conduits for the fluid hoses.

[Evengravy]

Good stuff John. I would have never thought of a concrete base but now that I see you mock up it makes perfect sense. It will also be much easier to surround the upper part of the walls with a side/back etc to catch coolant. Flood coolant it a really good thing to have on any machine cutting aluminum. It will save toy a ton of cutters! Good luck with the build.

Thanks, yes I think it should work out okay. I'm doing some calculations for mixtures at the moment in the hope to test the mix out this weekend in a very small batch. The only unknown is the shrinkage of the mortar for the walls/block but I'm hoping with a solid UHPC top any shrinkage should even out each side or be small. I'll try to get away with as little mortar thickness as possible but may test the Glass Fibre Reinforced mix as a mortar first to see if it bonds okay.

I'll upload sources etc when I get the initial test done.

[Evengravy]

That's a really interesting base design, I've been wondering about a concrete or block base myself in the future as a cost-effective and very rigid structure. I'll be interested to follow your progress.

If you plan to drain the coolant to a tank underneath then it may help to cast a short length or two of plastic drainpipe through the base to act as conduits for the fluid hoses.

We'll see how it works, I can't see why it will be a major issue unless shrinkage is significant over time, but I'll give it a bash. That's a good point on the drainage, I'll have to reiterate the CAD some more. The general idea was to have the coolant flood over both front and back and design a fibreglass chute to direct fluid back to the tank but that could also work, I'll have to have a think.

[Kitwn]

Just thinking out loud here John, but once you have the base epoxied up for leveling the rails you might as well use that surface as the base of your coolant-catching tray with a drain down to a storage tank below. Maybe build up the surface to direct the fluid to the drain. Rather like an Australian bathroom where the entire floor is built up round the outside to direct any spilled water into a central drain.

[Evengravy]

Just thinking out loud here John, but once you have the base epoxied up for leveling the rails you might as well use that surface as the base of your coolant-catching tray with a drain down to a storage tank below. Maybe build up the surface to direct the fluid to the drain. Rather like an Australian bathroom where the entire floor is built up round the outside to direct any spilled water into a central drain.

Hey, I think I'm following what you mean, that's not a bad idea actually, thanks. What I could do is run channels at the sides down into pipework and into the catchment tank from there, rather than fluid running over the front and back, and build a lip at the front and back for liquid catchment. That might make things a bit difficult loading and unloading stock though, particularly if stock is to overhand the Y (I do that sometimes).

The spindle only gets within 150mm or so of the alu sections which holds the rails left and right so there's plenty of space there for a liquid channel. I must draw up some options to gauge the simplest approach before starting.

[Evengravy]

Initial test mix is done, the testing here is designed to judge bonding strength when using a simple UHPGRFC mix as a replacement for mortar. This might be unnecessary but mixes like this are said to be good for 12,000psi or so, so I'd like to try it.


UHPGFRC mixes (in my very limited experience) are approached differently than regular cement type mixes, insofar as the cement to aggregate (sand) proportions are quite different. With regualar cement mixes you would generally be 1:3 cement to sand ration, here we are 1:1 roughly.

I have a simple mix and a more exotic one, the simple mix is approached here as I'll need to source some other part for the latter which I'll detail later.

Here we have:


28150

+200g Sharp Sand (quartz I believe) I live near several quarries and this stuff costs me roughly 25 euro per tonne, give or take, and makes great concrete.

+200g Portland Cement (regular cement). Generally this will be mixed with other elements such as 20% CSA or Silica Fume, I'll fill you in on that later, those can be hard to source.

+6g of Glass Fiber (From mbfg.co.uk added to the mix last, strand length is 13mm in this case, probably too long but it'll do the job)

+0.6g of Plasticiser (FLUP PCE 104 from moertelshop.de) I went with the manufacturers recommendation in this case of 3g/kg, it probably needs to be doubled in reality though, so I'll have to order another 1kg tub for the table to be sure.

Don't take anything I say as gospel, this is what works for me so far, I'm very new to this so take everything with a pinch of salt and as 'in testing'.
The mixing procedure used was pretty straightforward, mix sand and cement together dry first, add a small amount of water to make an overly dry mixture that is just starting to clump a little. Add super plasticiser, continue to mix for a minute or two (vigorously) and add more water to bring it to field capacity being careful not to overdo it. When the consistency was where I wanted it, I mixed in the glass fibers last.

I ended up using a little more water than I would have liked or predicted from my calculations (65g), but I think I'll double the plasticiser next time and test again to see if I can further reduce the water input.


28149
I used some small 'soap bar' bricks which are almost identical in nature to the larger versions to test it out, at the moment I'm mainly concerned with bonding strength if I use it as a mortar between the blocks. If it fails to bond, I plan to make another custom mortar mix using standard mortar mix with sand, cement lime and mortar plasticiser with some added glass fiber reinforcement to beef it up as I know that will stick well to the blocks.

One thing I've noted is that the mixture consistency becomes noticeably more 'dry' when adding the Glass Fiber, so I'll have to balance that out. Also the working time is short, around 20 minutes, so I'm going to have to be on the ball when doing this but I have a cement mixer so it should be doable.

The mixture is curing very fast, so in a day or two I should have an idea how the bonding is.

[Evengravy]

Some small progress.

After letting the GRFC mortar test cure for a week in situ in my workshop I inspected it for bonding, the very scientific method of slapping it with a hammer was employed. I was mostly concerned initially with whether it would bond significantly to each brick but I'm confident now that it will be perfectly fine in that regard. It's clearly much stronger than regular mortar and I'd expect it to shrink less and be stronger in compression. That's good enough for me to move on.

One thing I realised after purchasing the glass fibre and mixing up this test is that AR glass fibre is recommended as it is suitable for the extreme environment of usage in concrete (general E glass is not as it is attacked by the alkalinity). After reviewing the technical documents from the supplier though it seems that this isn't a major issue for me, luckily. The glass fibre I have purchased, whilst not AR glass, the type I have (E6-CR from Jushi Tech http://en.jushi.com/res/excel88fcd3bf87268a79.pdf) is tested and recommended in both high acid and highly alkaline situations so it should be fine. I note it here for others though, be careful, glass fibre is not all suitable here and go for something that is specified for concrete usage.

Here's a picture of the GFRC mortared test, being reused as a weight. This mortar is tough stuff. I'd like to do some more scientific tests before casting the main bed, I'll get the walls up firstly.

28207

I'm pretty slammed at work but had some time to layout the basic structural outline on my newly concreted floor. Nothing too complex, four bricks, some string and some careful corner to corner measurements to make sure it was square. One of the issues I was having with the concrete floor was it was endlessly dusty, no matter how much it was swept it kept producing more and more dust. I treated it once with a PVA based sealer and water and this has cut this down. I will aim to treat it once more before moving on with the block work next week. This should also help ease my mind around bonding to the floor, I was more than a little concerned that it wouldn't bond correctly due to the dust but I'm told this is a commonly used solution to aid bonding.

28208

Next step for me is to get my workload tied off and begin on this around the 10th.

I've ordered some new motor to ballscrew connectors some months back but they have yet to arrive, if anyone knows of a reliable UK source I'd appreciate it. I'm looking for something to handle 480oz/inch with minimal backlash. 8mm to 12mm. Without a lathe in situ I can't really modify what I have accurately and haven't as yet been able to fully test the new ballscrew install.

Anyway, that's it from me for now, more soon.

[Mister Mills]

Hey John,

I have 3No 8mm to 12mm couplers on the way. They should be with me in the next few days. I have no need for them as they were part of a package deal. I have upgraded my motors to Nema 34's and I have ordered the 12 x 14 Couplers that I need. You would be more than welcome to have the 8 x 12's for nada... Otherwise they will just sit in a box for eternity.

[Evengravy]

Hey John,

I have 3No 8mm to 12mm couplers on the way. They should be with me in the next few days. I have no need for them as they were part of a package deal. I have upgraded my motors to Nema 34's and I have ordered the 12 x 14 Couplers that I need. You would be more than welcome to have the 8 x 12's for nada... Otherwise they will just sit in a box for eternity.

Hey, that's very kind I appreciate that, I might take you up on that if you really don't mind. I've ordered two pairs actually but this whole pandemic seems to have put a spanner in the works to getting anything from China. I'm not in a mad panic but if there's nothing next week can I give you a shout? I'm happy to fire some money over for them.

[Mister Mills]

No problem John,

They are still on the way for me. Currently in Germany so I would expect them at some stage next week. Looks like I am going to be waiting for my new motors and drives a while more than I thought also due to C19. It's a pain in the ass. I'm drooling over that piece of level floor in your workshop. I'm building on a ski slope at the moment hahaha

[Clive S]

[QUOTE]I treated it once with a PVC based sealer and water and this has cut this down. I will aim to treat it once more before moving on with the block work next week. This should also help ease my mind around bonding to the floor, I was more than a little concerned that it wouldn't bond correctly due to the dust but I'm told this is a commonly used solution to aid bonding.
[/QUOTE

Looking good, I take you meant PVA or even better is SBR

[Evengravy]

[QUOTE]I treated it once with a PVC based sealer and water and this has cut this down. I will aim to treat it once more before moving on with the block work next week. This should also help ease my mind around bonding to the floor, I was more than a little concerned that it wouldn't bond correctly due to the dust but I'm told this is a commonly used solution to aid bonding.
[/QUOTE

Looking good, I take you meant PVA or even better is SBR

I did yea, thanks I didn't even notice that error (edited) it must have been the autocorrect. I'll look up SBR before beginning as I'd like to treat it again if they are compatible with one another. Thanks for the heads up.

[Evengravy]

Okay,

So a bit of progress. I slowly managed to get the walls built for this machine base. I ended up mixing small batches of my previous glass reinforced mortar mix, essentially just scaled up to larger batches and continued building one layer at a time, leaving a couple of hours between layers to allow mortar to go off slightly.

I ended up with 3.5 blocks long per layer and five layers high in total, to bring it roughly to waist level. I began by pre-cutting the granite/concrete blocks with a diamond disk on the 4" grinder and cut around the circumference in the centre, blocks were then easily separated into pretty clean halves through some brute force with a bolster and club hammer. It doesn't take a huge amount of effort in reality. I certainly won't win any awards for my block laying skills, but this mortar is more difficult to work with tidily than regular mortar due to the glass fibre input. It should be worth the effort I feel as it is tough stuff, so I don't see any issues with it. It should mostly be under compression here so hopefully shrinkage will be a non issue later. We'll see I suppose.

I tended to try and keep the mixture on the drier side. It is more difficult to manage that way but I feel this will be a good approach in terms of the overall strength. The water volumes are only a guideline really, in practice each batch will differ slightly so half is added, and gradually added there after to avoid making it too wet. With the super-plasticiser it's easy to overshoot the water addition and end up with a mix which has too much flowability.

283792838128380

I ended up leaving some space around the perimeter in order to facilitate access and to allow sufficient room for any future enclosure.

2838328382

Thanks Clive for the heads up on SBR, I've purchased a gallon of that to treat the floor and paint it later. I'll likely get the messy 'table-top' casting out of the way before I do that though. That's the next job on the list.

[Mister Mills]

Not long now John........... Looking pretty sturdy! I have been goosed for time myself but have started do do a few bits as stuff arrives. Delivery times are dreadful :(

[Evengravy]

Ah, it's slowly taking shape. Thanks Liam. Can't imagine it's going to fall over any time soon so I'm pretty happy it worked out. Next job is sourcing some plywood and steel rebar so I can prep a mould to cast the top. All being well I hope to get started on that this weekend if I can get everything I need. I'll need to buy some more cement and check on the status of my plasticiser order, it should be here soon.

I ended up ordering some more motor couplers from a European source this time, third time I've ordered these similar parts now which has been both annoying and unnecessarily costly but I don't hold much hope of anything I've ordered from China ever arriving to my door. I'm a cynic in general but part of my thinks this whole thing is really about putting a pin in Chinese manufacturing dominance. Maybe that's a good thing for us and this particular interest but anyway, maybe I'm just getting too cynical as I get older.

Anyway, if this order doesn't come through I'll definitely hit you up about those couplers again if they come through to you. I'm hoping mine will be here in the next few days so I can finally test the ballscrew install.

[Mister Mills]

Hi John, My order arrived last week. A month late!!!!!!!!!!!! I have 3 pcs of spare 8 to 12 mm couplers. PM me your details and I will pop them in the post if you still need them.

My motors and atc wont be here for another 3 weeks at least :( so I well know how much of a pain it is to be waiting for stuff to arrive.

I agree with the whole cynical outlook. It's true haha. Same myself.

[Evengravy]

Thanks, I'll give these ones a few days and if nothing by the end of the week I'll give you a shout. The first order I placed was at the start of April so I doubt they're disappeared somewhere at this stage. Thanks a lot though!

[Evengravy]

More progress made over the weekend. I didn't quite get to the table top pouring but have the majority of the prep done. I also received the ballscrew to motor mounts (finally) so was able to install and test those. No issues so far, noise is greatly reduced and I have it running up to 6 meters/min currently. Beyond that I get stalling but may be able to tweak it a little more later.

Next step in the table build was to build a mould for the concrete top which will be cast soon. I'm awaiting some more super plasticiser but that should be here soon. Calculating the approx weight it will be in the order of 400KG when cast, so my first concern was bracing the open section well.

28433

I chose to use studding which was left over from the building of the workshop in which this machine resides. I made up three sets of braces one for each side and a beefier version for the middle section. I chose to do this with the consideration that it needs to be knocked out from underneath later: I want it to be solid but also removable after the fact. There were some additional braces between these sections that isn't shown here.

On top of this some 11mm ply was set and sealed with silicone between joints. I chose to have the ply a little lower than flush with the walls, this might be difficult to see in the photographs but the idea is that the concrete pour will step down 10mm (vertical) over the end of the block walls before meeting the ply on the inner side of the blocks, creating a key of sorts. I thought this would add some additional strength.

28434

Toward the front I added some 34mm (I think) plastic conduit so I can cleanly route a few wires from left to right (front left Y motor and home/limit switch). This is cut to be a tight push fit against the ply walls, which are fixed to the walls with masonry nails and a 600Kg ratchet strap is added for additional support from bowing. I also ran some lengths of string through the conduit/pipe and fed these through the ply walls at each end such that I can easily pull my wiring through post casting later.

28435

All the wooden surfaces that will be removed were given a couple of coats of beeswax to function as a release agent to aid removal post concrete pour.

28436

Lastly, high tensile steel rebar was added in an overlapping cross pattern. I chose to tack weld these together with the ARC welder and prop them up towards centre (ish) with some rock that can stay in situ as I cast.

28437

I've got a few little gaps here and there to fill with silicone but as soon as the plasticiser arrives I should be good to go on the casting of the top section. I roped my brother in law to help out, likely it will be Saturday next when I get started on that bit.

[Evengravy]

Not sure if anyone's still following along, but I'll pop stuff in here regardless for sake of completion.

Concrete top has been poured. All in all it wasn't very difficult with the help of my brother in law and a cement mixer. I ended up pouring just over 500KG of my initial glass fibre mix scaled up. I was a bit nervous that the mould wouldn't hold up to the pressure/weight, but it seems to be okay (touch wood).

Just for reference, the mixer held:

50KG of sand
50KG cement
to this I added:
150g of super plasticiser
1.5KG of glass fibre
circa 8L of water, added gradually to play the mix by ear a little.

I ended up with five of these mixes in total in this top section. I don't envy the person that has to remove this, and hope that it isn't me....

2848628487

I've covered this in plastic sheet (not shown) to prevent it from drying out too quickly so that it can cure properly and I'm going to have to give this a mould at least a week before touching it. What I'll likely do is leave the mould sides in place and cast the epoxy top before removing the sides. I do need to do some careful figuring out to maximise the use of my z travel though. There may be a case where I'll have to increase the height of the Y rails to this top to give me sufficient clearance to fit my vice etc. but that is to be determined and most likely unnecessary.

[Kitwn]

I don't envy the person that has to remove this, and hope that it isn't me....


Don't worry. The next owners of the house can turn it into a poole table :excitement:

Looking forward to seeing the final machine and how you make the gantry to match the base.

[Evengravy]

Don't worry. The next owners of the house can turn it into a poole table :excitement:

Looking forward to seeing the final machine and how you make the gantry to match the base.

That's a good point! Next up epoxy layer, I'm still looking around for materials atm. If anyone has any heads up on a low viscosity resin I'd appreciate it. I used one years ago on the Momus CNC base but completely forget the manufacturers name. I'm sure I'll find something though.

[Clive S]

That's a good point! Next up epoxy layer, I'm still looking around for materials atm. If anyone has any heads up on a low viscosity resin I'd appreciate it. I used one years ago on the Momus CNC base but completely forget the manufacturers name. I'm sure I'll find something though.

Wests system https://www.westsystem.com/the-105-system/epoxy-resins-hardeners/

[Evengravy]

Thanks Clive, I'll take a look at that one.

[Mister Mills]

Coming along nicely John. I really looking forward to seeing her up and running. :)

[Doddy]

.

[Evengravy]

Me too! It's been a long time but we'll get there. I'm actually pretty glad I did the wiring aspect earlier, I think that the thought of tackling that again right now would be overbearing.

The top is drying well but not as flat as I would have liked, there is an area in the centre where the concrete is 5mm or so lower than the rest, I'm not entirely sure if it was bowing under the weight or just due to some factor related to drying but it's not the end of the world.

I have sourced the resin, it's here with me, I ended up going to a crowd in Northern Ireland but I'll fill in details later around that.

Prior to pouring the top I intend to flatten it as much as possible, reason being the resin isn't cheap and I want to have a nice even layer across the surface. To that end I have ordered a Diamond Cup for grinding/polishing concrete that will fit to my 4" grinder. It's on the way so should be here soon. It shouldn't take too much effort to knock down the high spots of the top with this tool to prepare it for the epoxy layer. I might combine the grinding with some filling of the very low area with epoxy/granite to minimise grinding but we'll see how we go.

I have a good respirator and will likely try to work wet with the grinding if possible, I know silicosis from breathing concrete dust is nothing to toy with. I really would have liked to have avoided grinding the top but those are the breaks, it will all work out I'm sure. As soon as the grinding disk arrives I'll be getting started.

One thing I'm a little concerned with in relation to the epoxy pour is the temperature in the workshop, looking at the resin data sheet we need to stay above 15 deg. C to avoid issues. I'll have to plan ahead with weather and some external heating source (heat lamp maybe) but I'll have a think, I'm sure we can come up with something.

[mekanik]

You might be better off trying self leveling floor compound, it runs like water.
Regards
Mike

[Evengravy]

You might be better off trying self leveling floor compound, it runs like water.
Regards
Mike

Thanks Mike, I'm kind of committed to this way of approaching it now as I have all the bits and bobs ordered. The resin I have is really low viscosity so it's really just a case of roughly evening the concrete surface before hand, shouldn't be a massive chore but without doing so the majority of the resin would sit in the centre leaving an overly thin layer at the sides where the rails will sit. It's not far off as it is though, about 5mm in the worst spots so it'll be fine with some gentle massaging. Appreciate the point though, I might source some of that to level any low spots on the floor of the space itself before painting it.

[Doddy]

Are you comfortable that the concrete will have cured sufficiently, and is not gassing, prior to pouring the epoxy? I'm not going off anything other than tales that concrete takes a lot longer than some might expect.

[Evengravy]

Hey. I'm not 100% on the curing time but really just playing that by ear at the moment. The grinding process on the top is complete now and that proxess was useful in order to gauge the state of curing. Really messy job but was able to get it within a few mm across the surface. I plan to pour it toward this weekend which should be over three weeks since it was poured for it to cure so my hope is that will be sufficient. Once the epoxy is on top my hope is that any remaining drying will be allowed for by the sides and underneath that will still be open to the air. I'm aware I'm experimenting here though, so it could prove to be an issue, but ill give it a bash to see how it goes. One thing I'm certain of at this point is the strength is there, even with a diamond cup on the grinder I was barely able to grind away the top, slow going. Whether that means a lot in relation to off gasing is beyond my knowledge to be honest

[Evengravy]

Today is epoxy pour day (number one)

I decided to pour the top in two stages. Reason being I wanted to ensure the first layer goes on without a hitch and apparently it's good practice to do it this way.

Step one: I went through the form and sealed all the edges with a small amount of epoxy, allowing it to fully cure before keying it in with some sandpaper. Reasoning here was to be 100% sure that my expensive resin wasn't going to leak down the side of the form. I'm glad I did this first as I was able to spot a few problem areas that needed a couple of coats to seal.

2860528607

Step two: I was really concerned about the temperature inhibiting curing. The resin I have is documented to work best above 20 degrees C but will cause issues if the temp drops below 15 degrees. Daytime temps at the moment are above 15 so that isn't an issue but nighttime temps are dropping to 11 or so.

What I ended up doing is buying a garage fan heater and a 13A plug thermostat. By sealing the underneath section of the form with some ply and blowing the hot air in, the idea is to heat the mass of the concrete base and prevent any issues. I'm told the air temperature doesn't matter so much so long as the surface temp is above the threshold. This seemed like the most efficient way of getting the heat where it's needed without breaking the bank on electricity. We'll see if it works.

2860628604

I'm currently in and out of the workshop to pop any troublesome bubbles with a hot air gun, so far so good anyway.

Best, John

[Evengravy]

After 48 hours to cure the first layer of resin is on and cured.

I'm pretty happy with the result so far. One thing to note is that the concrete, being porous, caused many bubbles during the set up. The resin I have has a very long pot life: >8 hours @ 25C so I was in and out of the workshop every 30 minutes during that period to pop any bubbles as resin sank into the pores. It wasn't much hassle to be honest, the heat gun takes the bubbles out in a flash with a quick pass. Incidentally I should mention I have an adjustable heat gun (designed for reworking PCB's) so that was set to the lowest heat setting in order to avoid overheating the resin.

After cure I'm quite happy with the results, it's self levelled very well indeed. I have a sensitive Starett No 98 machinists level and it's no more that fractions of a mm off in the worst spots. I'm also glad I chose to do it in two stages, the porous nature of the concrete left some areas with creators as the resin sank into pores during the end of the pot life leaving small craters. At this stage it will be completely watertight so the next layer should even that out nicely.

I plan to leave it to cure for another day or so as I get on with other things following this I'll key it in lightly with some sandpaper, hoover up the dust and pour the second layer in the same way.

I've been thinking ahead to mounting the machine and have some materials on order for that which should be here soon.

286212862228623

[Evengravy]

Second layer of epoxy has been applied and cured. The heated arrangement I had worked out really well actually and kept the entire thing over 20 degree C for the period. I ended up pouring a total of 14.5KG of resin (I know!) on this and I'm glad it worked out well.

As far as I can measure it's completely level. I actually spotted a small error in the calibration of my Starett 98 machinists level when rotating it 180 degrees in the same exact spot. These can be adjusted at the side reasonably easily and I was able to dial it in after a few tries. All in all, the two stage approach to epoxy was the right choice in concrete for sure.

So today I removed the sides of my form and chamfered any sharp edges of the epoxy by scraping with a stanley razor blade, much like you would using a cabinet scraper on wood. The sides came away really easily and the beeswax did a great job as a mould release.

I'm planning to leave the underside supports in place for now, I'm in no panic to remove them and I'm a little cautious to remove them until I'm sure the top is 100% dry.

Hopefully we won't see any movement of the top over time, time will tell.

I've ordered some bits to start fixing the machine to the top. Plans are pretty simple here: angle iron at each side of the Y rails bolted to the concrete with Hurricane Bolts/Anchors then bolting the profile to the angle with many 6mm bolts.

I'll fill in the remaining details as and when I get to that though.

286512865228653

[Evengravy]

Progress over the last few weeks has been okay, I've had a bit of time here and there to get things together.

28793 28795

I've fitted the machine to the epoxied concrete base now, I ended up using 40mm steel angle (5mm thick) on both sides of the Y rails (inside and outside). These angles were drilled to accept 6mm bolts into the aluminium profiles every 80mm and were fitted to the base by drilling with an 8mm masonry bit to 80mm depth and bolted down with 24 of 75mm x 8mm Thunder Masonry bolts/anchors each side (https://bit.ly/32X7aeP). I chose these anchors as they are said not to stress the concrete when fixing, can be removed are strong and well plated against rust. I don't know what the pull out specs are for these but I'm guessing they'll be more than good enough for this application.

This thing seems really solid now. The overall job of fixing the machine to the base took a little longer than I predicted as there was an alignment issue with the machine causing it to be 1mm wider at the rear than the front (measured between Y rails). This was caused by a less than perfect alignment of the plates which connect the gantry at each side to the Y rail carriages. To rectify I had to remove them, probe them in on my other cnc an re-machine the bolt holes. I replaced M5 hardware at that time with M6 and at the same time and counter sunk the socket head cap screws. After some careful adjustment and refitting it seems to be rectified.

The metal work on the angle iron/steel took longer than I expected it to, I got it done with some persistence in the end anyway. Steel angles are all painted now (I ran out of black) to protect from rust a little but I have to order in some other paint for any remaining steel parts (gantry to ballscrew brackets).

Next up is rigging up the wiring and figuring out the cable chain management on the Y axis, I should make some progress on that this weekend (welding up some brackets at least).

28792 28794 28796

At that point I should be able to get it moving around so I'm looking forward to that. I have a few more stages to go though: workpiece hold down, vices and I'm planning on epoxy/granite filling the Y rails and potentially the gantry too.

Before I fill the gantry I intend to strap an equivalent weight to the gantry (I calculate an additional 10Kg approx) and run some performance tests to see if my rapids are altered significantly by the extra weight. The gantry isn't excessively long so I'm hoping the performance will be okay.

[pippin88]

Is your steel angle truly 90 degrees? Often they are not and that could cause issues for your alignment.

[JAZZCNC]

Why did you epoxy the whole base.? You will need a base board or fixture plate and that will need to be surfaced so it seems like a lot of wasted epoxy to me.?

To me sitting the rails on profile to raise the gantry defeats the point of the concrete base as it introduces a weak point which could have been avoided by creating raised sides by shuttering up and filling with concrete, then epoxying just the raised surface. You could have then fit tooling plate to the surface for the rails to fasten too, it could also have been used to fine tune out any errors.

But I suppose the beauty of concrete is that you could still do something like this if you wanted without too much hassle.

[Evengravy]

Hey jazz. I mainly did it this way to have a nice level surface to measure from to align x travel but you're right, it's not the most efficient use of epoxy for sure.

I've spent quite a bit of time aligning everything.

Front to back I'm within 20 micron on the left Y rail (measured rail block to epoxy) in the right side within 50 micron. If I were to do it again I'd cast the y rails as raised concrete sections as you say jazz. It'd be closer then. The error is coming almost exclusively from the profile, but, I'm happy enough. I'm going to try hand scrape the alu Y rail on the right hand side to get it down to within 20 micron. It's bowed up 50 micron in the middle of travel. Not a massive amount but can be improved a little. These figures are front to back, in smaller areas the accuracy improves significantly.

On the X axis I'm within +/- 5 micron left to right so all good there. At those levels I'm reaching the limits of my measurement tools/technique.

I'm planning to drop a cast Alu plate on here fixed with threaded anchors soon ish. But I want to get the alignment as close as possible before I do that. That's really why I epoxied the entire top. Maybe not the ideal (cost effective) way but it gives me peace of mind when measuring. The project is on hold atm as I get the electricity supply sorted so will pick it up again soon. I'll really need to insulate the space too given the weather has gotten so wet here.

Comments are bang on though, could be done more efficiently and if I were to build again I'd cast the Y rails from concrete. Basically I had these already done and the thing running before designing the base so didn't want to start again redesigning bits, but so far it's working out alright. It's a learning experience so I expected that to be honest.

[Evengravy]

Is your steel angle truly 90 degrees? Often they are not and that could cause issues for your alignment.

They're aren't exactly 90 degrees as you'd expect, you're right, but since I'm bracing both sides of the rail there is scope to adjust allignment each side, so it isn't really an issue in my case. Alignment isn't an issue so far, bar the time required to dial it all in. Chasing microns is slow going.

[Evengravy]

@Jazz

Thinking ahead to fixture plate. What grades would you recommend for cast plate? It's not something I've purchased personally and suppliers on this side of the water are much more limited than the UK. Am I correct in thinking it would be cast machined 5083? something like this:

https://www.impactirl.ie/product/cast-machined-aluminium-plates-impact-500-en-aw-5083/

[Evengravy]

Hey,

It's been a while since I posted here, hope everyone is keeping okay. I haven't forgotten about the build and have been taking baby steps whilst I try to get someone to run an armoured electricity supply to the 'shed'. I insulated the place, painted the floor, fitted and painted the walls, build a solid bench and acquired two small lathes in the mean time (one is a emco cnc lathe with tool turret that I have been converting to Mesa/Linuxcnc at the same time, the other is a 'Chinese' manual lathe that I've been modifying with epoxy granite and some tapered gibs.)

30000


29999


Anyway, back to the router; I have sourced a cast aluminium plate (15mm 5083) and am about ready to fix it permanently to the bed. For fixings, I intend to use the same hurricane bolts I used for the Y rails, that's all simple enough.

Before I started milling this plate I wanted to get everyones take on the bolt holes for fixtures. What spacing would you recommend? I did a design on 30mm spacings (similar to the Saunders Machine Works designs) but it came in at 1700 holes to drill and treadmill so I'm not sure I could manage that before I have the flood cooling system in.

I went and purchased some good quality carbide tooling though, for the drilling I went to cutwel and got a 5mm alu-specific parabolic carbide drill (made by YG-1 I believe) and thought that, maybe, with the right pecking cycle and some care I might be able to avoid gumming up the cutter with manually applied cutting fluid.

What's your take on M6 bolt holes spacings on a plate like this? Also, anyone have advice on speeds/feeds/strategies for dry drilling 5083 with parabolic carbide tooling? I haven't thread milled before so that's going to be fun.

[Evengravy]

If anyone was wondering if this machine was ever finished, it was, and is in use daily. Really happy with how it turns out it chews through aluminium and is very accurate. Would I recommend glass reinforced concrete? Yes, its stable after the years. But, probably not the most efficient way to build a router. For what its worth, time well spent in my book. Sorry for digging up the old build log but, I wanted to finish the thread off with some long range thought for anyone finding it. Im making a fixture plate at the minute, I will try to gather some pics later.

[Evengravy]

I ended up fitting a pre-ground aluminium tooling plate, self drilling and tapping it as a workplate surface. At the moment I am doing a lot of work with Aluminium so am making other multi-op fixtures that sit on top of this. If I need to work with wood, I simply bolt a sacrificial wood top to this alu plate. I still have some improvements to make here and there but works just fine. I am very glad I wend with quality rails and screws it is very accurate and repeatable. Pretty solid setup, but that could always be better. I'm normally milling at between 1500mm/min and 4000mm/min depending on the end mill in aluminium. I really like the DLC coated end mills for price vs performance ratio but have some 'better' (re: more expensive) YG1 tools for certain situations that call for them. Those are seriously impressive!. Most jobs don't justify the price difference, though. Anyway, here's a picture of the working surface plate with a fixture in the works for a three Op part. I designed some modular Mitee-bite fixtures that come in very handy, too. You can see those in the shot. Sort of like a modular vice idea.

32000

32001

Not sure why the pictures get rotated 90 degree but never-mind, hopefully you can see what's happening okay.

P.S. I don't normally have the spindle extended that far from the spindle clamp, it was just necessary do drop it lower for this job. What I will probably do soon is make another clamp and make a new backing plate to the Z axis to stiffen it up some more at long extensions. Normally my work is much closer to the gantry than it is here. Anyway, works fine for now.

I haven't found the time to address the flood coolant, yet, but the design is such that it would be relativley easy to implement. The tooling I'm using currently doesn't really need it so I may address that if/when I need to work with harder materials. Currently, it's mostly aluminium, some carbon fibre and some wood that goes on here. I have a manual mill for any steel and can do it on here too with the right tooling.

If anyone is wondering, I also got the CNC lathe running on Linuxcnc including the toolchanger. I went with a MESA 7i96 and its a great little machine for the size.