El Beast - Initial design phase, comments and critique welcomed!

[Desertboy]

Cheers, though its a toy compared to some of the beasts on here (looking at you chaz!). Don't have any drawings, what are you after specifically?

Rob - ta!

I noticed today that I've had a bit of a design snafu. My motor mount model was mislabelled - the model is a mb10-c and what I have are mb12-c.

Thankfully I caught it before I made the x or z parts, but it means my ball screws are 3mm too low at the motor end as the mb12-c is a little taller.

So options are:

- Buy 2x MB10-C (bad... about £60 each and it isn't great for consistency)

- Re-machine a new front face plate with the ball screw bearings moved down by 3mm (I'll have to order more aluminium... not great but cheaper than two new motor mounts). Will also require a small design tweak to the ballnut mounts.

- skim the motor mount plates down by 3mm. Downside is they were only 10mm to start with so would be getting a bit skinny.

I'm leaning towards option 2. Really annoying - good lesson to verify all models are correctly labelled before using them! Could be worse, at least I caught it before it caused problems on all the axis.

Feel for you, I almost made a similar mistake I just assumed all my Hiwin's were HGH carriages but some were HGL which are 4mm shorter. I almost missed it it's only because I put all my Hiwin's together that I noticed there was a height difference and so actually checked with a steel ruler and discovered the recovered Hiwin's were HGL.

Really shows that measuring everything in the real world to make sure you cad models are correct and the right ones is just a good idea.

Can you get tooling plate locally? Where I am there are a few suppliers doing tooling plate they are more expensive than aluminium warehouse but cheaper on small orders because there's no delivery charge.

[Zeeflyboy]

Not aware of any sadly, but never mind - can order some more stuff next week from alu warehouse.

So given some doubts raised about the accuracy of the shoulder bolts as a way of aligning the mounting plates, I thought I had better check what I was getting. Given the gantry arms will sit on a saddle plate, these mounting plates do dictate the squareness of the Z to the Y axis, so the closer they are to perfect the easier time I will have making sure the gantry is squared up.

I figured this was a good method of checking - in theory if the mounting plate is perfectly level then the reading on the indicator should stay at zero as the plate moves from front to back (indicator is metric, 0.000mm)







Magically, this side was 0.000mm both at front and back, the middle showed -0.006mm which I would guess is due to non-straightness in my current machine's X-axis gantry meaning that there is a slight bow in the top of the plate, but 0.006mm is only slightly larger than the some bacterium is long so I'll forgive it that!

Sadly, my joy at that stunning level of alignment was short lived - the other side showed one end about 0.072mm out between front and back. Not that bad over a 200mm or so span, but not really at the level of what I'm after.

So two options - either I just skim down one of the shoulder bolts and surgically tap the mount into alignment, or I make life difficult for myself and add a shoulder for the carriages.

I decided to try the shoulder method, since i can easily change back and forth between either method then depending on the results of my experiments. The only catch is that since my current machine isn't exactly perfectly accurate, I was largely relying on the flatness of the eco-cast plate to give the required alignment and tolerances... that means machining a pocket to then give the carriages a shoulder to rest on is not ideal as I've then lost the flatness of the plate and am back to relying on my machine's tolerances.

I decided to try machining a slot, then making a counterpart to go in there which would provide an alignment rail whilst still having the carriage mount to the ground surface of the plate.

This is the design:




Doing it retrospectively, I would have machined the whole thing back side first, then flipped over for pocketing the heads of the screws... that would have ensured the absolute best possible alignment between slot and mount plate. Seeing as that option is no longer possible, I used some 6mm H7 dowel pins and pre drilled a pattern into the bed so that the plates are aligned as best as I can manage to the machine.






Then machined the inserts:






They fit very snugly, can't feel any play even without the screws in place.







I then re-ran my indicator test, and while an improvement over the poorer side of 0.072mm, when compared to the perfect side it got significantly worse - it is approximately 0.020mm higher at the back than the front. It is fairly easy to knock it into shape to within around 0.004mm between front and back by loosening the bolts a touch and giving a little tap. I may try just using a little shim material to give a more precise fit. I think aiming for anything more precise than that really is just chasing my own tail since the first test showed the surface isn't flat to more than 0.006mm anyway.

[EmilW]

Really shows that measuring everything in the real world to make sure you cad models are correct and the right ones is just a good idea.

Thats some realy good and important advice ^^.
Always make sure the real world is up to spec and the reverse :p

Made that error many times myself.

[routercnc]

Hi Zeeflyboy
The check you doing at the top of post 143 doesn't guarantee that the side plates are parallel to the bed. They might be pointing up or down hill. As you are measuring at one point you won't see this.
So repeat but this time keep the side plate fixed and slide the mag base instead. If that is also ~0 then the two parts are parallel.

[Zeeflyboy]

Indeed, it doesn't guarantee alignment to the bed, but to the rail/carriages... which are aligned separately to the bed. If the plate was at an angle to the rail/carriages the readings would change as it moves past - I can actually demonstrate that by inducing an angle.

If measuring to the bed then yes, it needs another point of reference... but moving the indicator is enough to induce a small change just when it comes down into contact with a different part of the plate so that's not quite so easily done. I figure it's easiest for me to align the rails to the bed, then align the side mounts to the rails.

[1Jumper10]

I've never anodized before but if it's important to agitate the solution of sulfuric acid, I would be reluctant to use air to do it. I wouldn't want to create any kind of vapor to begin with and then I wouldn't want it being carried out into immediate area where it could cause problems. Breathing and corrosion and such. Is there some reason you couldn't use a motor driven polyethylene agitator? Just a thought...

Sent from my Pixel XL using Tapatalk

[Clive S]

I've never anodized before but if it's important to agitate the solution of sulfuric acid, I would be reluctant to use air to do it. I wouldn't want to create any kind of vapor to begin with and then I wouldn't want it being carried out into immediate area where it could cause problems. Breathing and corrosion and such. Is there some reason you couldn't use a motor driven polyethylene agitator? Just a thought...

Sent from my Pixel XL using Tapatalk

I have never done any work with sulphuric acid either but the resident forum member has done so much in his life so he has to be the go to person for all advise on all matters. So it MUST be the correct way to agitate the acid. (I don't think) perhaps with an 80 cfm compressor.

A bit of reading here http://www.tsh-technic.com/sites/def...ons-update.pdf

[Boyan Silyavski]

I have never done any work with sulphuric acid either but the resident forum member has done so much in his life so he has to be the go to person for all advise on all matters. So it MUST be the correct way to agitate the acid. (I don't think) perhaps with an 80 cfm compressor.

These kind of remarks are absolutely unnecessary for the good tone here in the forum. I feel sorry when seeing such unfounded attitude.



I did the aeration using a simple silicone tube which i pinched many times with needle having previously sealed its end. Its important for the reason of evacuating the air bubbles that form when anodizing and moving the water to equalize temperature faster. The tube was connected to small silent air compressor with regulator , but any compressor from fridge would do, in that case a way should be found to vent out the unnecessary pressure build up .

[Desertboy]

Doesn't seem that complicated to anodize at home

https://www.youtube.com/watch?v=GccTVfMiIIo

http://astro.neutral.org/anodise5.shtml

I'm sure you know all this already but I find it interesting ;)

If you need to keep the solution agitated a small maxijet water pump would probably be fine they are designed for semi acidic environments I have a load of spare working pumps you can have one for free if you want and if it melts who cares. But leave it a small test jar first over night I think lol be annoying to destroy your acid with melted plastic.

[Zeeflyboy]

I assumed the same, ultimately there is a lack of stiffness and some small amount of backlash which would explain a small amount of surface imperfection. Its pretty much just a cosmetic issue and quickly polishes out.


Router - yes it would no doubt benefit from being a single piece but its too large for my machine to handle. It might be an upgrade that the new machine can make for itself later on.