PDA

View Full Version : Materials used in CNC build projects



GeorgeD
16-07-2010, 11:02 PM
So aluminium and steel work seems the norm for really major sized CNC machines.

Next is wood for the hobbyist machines of a smaller nature...what about Perspex/Plexiglass for the Gantry and y axis? is this material taboo for the smaller desktop build?

I'm talking 20mm to 25mm for the gantry and Y axis components.

Cheers. :smile:

irving2008
16-07-2010, 11:32 PM
its been used, there are some examples on small mills. But perspex has a very low modulus of elasticity so is approx 1/3 - 1/2 the rigidity of wood and 1/15 - 1/25 that of aluminium so has to be that much thicker and it tends to crack

GeorgeD
17-07-2010, 07:03 AM
Hi Irving

Yes it does have a tendancy to crack...but 24mm?,you'd have to give it a hefty wack with hammer to crack it.:wink:

I have some here,10mmI think? it bends but it won't crack don't know what grade type it is?

These guys used 1/2" acrylic thats a real nono,looks fantastic though.

http://rainbowlazer.com/3d/rhino/diy-3-axis-cnc-millenhanced-machine-controller/

irving2008
17-07-2010, 10:42 AM
perspex crack at stress points... it is striong in sheet form as you say, but drill a hole in it and you get cracks radiating from the hole, just as you do with any material, but perspex, as I understand it, cracks easily once its solidity is compromised

Ross77
17-07-2010, 11:36 AM
Don't forget that although different materials have different physical strengths and as such they also need to be used in way that best utilizes them, just swapping perspex for Ali in an design that was intended for Ali is likely to fail.

Metal has to be bolted (large parts to allow this) as welding can distort the shape but the advantage of acrylic is that it can be glued, (therefore design a monocoque structure) If you can machine small rebates for thin plates to sit in then and really triangulate the beam I would have thought that help reduce the problem with localized stress points that Irving pointed out. Hope this makes sense, I'm not very good at explaining things .

there is also the option of creating a composite beam. acrylic structure with thin sheet metal like a flitch beam or just adding steel rods to the tension side like a concrete lintel.

I need to brush up on plastics, not really used them, is there a difference between Acrylic and perspex?

Of course This is in the context of a small desktop machine:tongue:

GeorgeD
17-07-2010, 12:03 PM
Ok,I've got 3 sheets of 500x500x24mm how about if I was to weld it together and get some 1 or 2mm plate aluminium off ebay(quite cheap and I don't need a lot) and use 1/2" M6 seltappers to screw the plate to the sides of the gantry,that should strengthen it a bit.

Will use either Superglue orAraldite or even Dichloromethane ifI can get my hands on it?

florins
17-07-2010, 02:02 PM
...Will use either Superglue orAraldite or even Dichloromethane ifI can get my hands on it?

Hi George,

Everybody is trying to tell you "don't do it", but it seems that you need to try it yourself in order to believe it :). From my experience, drilled, cut and/or glued perspex is cracking like hell when mechanically stressed or vibrated (wich is exactly the case here). If you would use polycarbonate, there would be a different story, but still not matching the aluminium or even wood build. Now, if your machine will be very lightly used, it might work.

However, we all can learn, that's why we are on the forum, so if you do it, we might learn something...

All the best,
Florin

irving2008
17-07-2010, 02:10 PM
Well its an interesting idea George... though you might want to look at the weight of that by the time you're finished compared with other materials. Taking aluminium as 1, mdf is 2 and Perspex 2.5 in rigidity ratio for a square section beam, the weights are 1, 1.5, 2.8 in ratio, i.e. to replace a 10mm sq ali bar in perspex you'd need one 25mm sq and it would weigh nearly 3 times more...

How will it work in practice? No idea, never used plastics that way.

GeorgeD
17-07-2010, 02:32 PM
If the machine is vibrating,Florin then there's something amiss. :naughty:

florins
17-07-2010, 03:23 PM
If the machine is vibrating,Florin then there's something amiss. :naughty:

Exactly, the sturdiness is not there if the machine is vibrating :). With perspex you will have them both: lack of sturdiness (this means vibrations), and cracking on vibrations - perfect match :). Try to read what irving2008 is saying as well, you might understand why everybody is working with, at least, MDF.

But that's my opinion and I can be wrong. Not trying to discourage you, each of us needs to make his own mistakes, that's the way to learn and progress. Myself I am coming on the forum to learn a bit easier, and I am lucky to find here ways to not repeat other's mistakes. I will be among the first to congratulate you if I am wrong, but meanwhile, if you ask for advice, I am posting my opinions.

All the best,

Florin

GeorgeD
17-07-2010, 03:36 PM
Ok,I'm going to attempt it...only because I am eventually going to build the CNC in Allum but that will be sometime as I'm aquiring the gear to make my own Allum melting furnace with a view to casting the parts(can'twait to do this,then its party time. :beer::toot:
Slow in aquiring the parts though,only got the propane bottle,3 kilos of Allum,Prpane Gas Torch,Welders helmet.

Need an Apron,Gloves and crickters leg pads and a crucible. :heehee:

ptjw7uk
17-07-2010, 04:35 PM
I would look to get some large pieces of Ally to melt as the smaller the pieces the more oxide you will produce.
Ideally you need a reducing atmosphere for ally not easy with propane as you need a lot of air to get the temperature.
I would like to try but not enough garden(boss has too many plants)

peter

Ross77
17-07-2010, 07:44 PM
OK,I'm going to attempt it...only because I am eventually going to build the CNC in Allum but that will be sometime as I'm aquiring the gear to make my


Good on you. I wasn't trying to discourage you, just offering some options. If the usage and design can be matched then I don't see why It wont work. I just hope you arnt expecting this to mill the Ali castings you will produce:eek:

Start a build log stating exactly what size cutting area, type of spindle, motors and material to be cut and the materials you have already got.

I'm thinking something like A4, A3 max with a high speed dremmel running on metric studding for cutting foam blanks,pcbs etc......



Exactly, the sturdiness is not there if the machine is vibrating :). With perspex you will have them both: lack of sturdiness (this means vibrations), and cracking on vibrations - perfect match :).


I would tend to disagree with that :naughty:(not like me I know) it is a stiff machine that will vibrate, less sturdy machines will move and damp the vibration. The vibrations are caused by the machine in motion or the spindle and cutting tool, so if gentle accel , decel and a smooth spindle are used then it should be OK no?:smile:

florins
17-07-2010, 09:04 PM
...it is a stiff machine that will vibrate, less sturdy machines will move and damp the vibration. The vibrations are caused by the machine in motion or the spindle and cutting tool, so if gentle accel , decel and a smooth spindle are used then it should be OK no?:smile:

Hi Ross77,

My understanding is like this: the lower the stiffness (I mean more elastic), the lower is the resonance frequency for a given assembly. Rising the resonance frequency over the mechanical frequencies generated by the movements and the spindle makes the assembly less prone to vibrations - that's why everybody is trying to make the machines sturdier for a given mass.

It's true, you can absorb the vibrations with an absorbent material - like rubber, or, in this case, perspex :). It's true, as well, that you can reduce the accelerations (accel & decel) and the feed rate until a given structure will not vibrate - if you are lucky enough to not obtain resonance... but, is this practical?

Now, I am sure that George's machine will work somehow, but in my opinion, the pieces or perspex he has would work much better as material for some nice products after the machine is ready :).

Just my 2p.

Florin

P.S. Absorbing the vibrations in perspex I believe is exactly what's cracking it, and it starts where the micro-fracture points are induced by mechanical processing. It happened to me before...

routercnc
17-07-2010, 09:36 PM
Sorry Ross, I'm going to disagree with your disagreement :exclaim:

I think that a stiffer machine will vibrate less than a weaker machine. Why? Well the first explanation is Hooke's law:
force = stiffness x displacement
or
displacement = force / stiffness

For a given force (from the cutter), a higher stiffness gives a lower displacement.

But we are also talking about dynamic stiffness and need to consider resonances. Undamped natural frequency = SQRT (stiffness / mass). A stiffer system will have a higher natural frequency, and in turn these have lower amplitudes. I believe this is to do with higher frequencies having more energy in the wave. Since you are putting the same energy into the system (from the cutter), the higher frequencies must vibrate at a lower amplitude to keep the 'area under the graph' the same in both cases.

Higher frequencies can be confused with 'more vibration' because to the touch they will tend to give more of a tingle as the frequency increases, but if you were to measure the displacement it would be lower. Run a sine sweep through a loudspeaker and watch the cone. Low frequencies you will see it move - at high frequencies it will virtually be stationary. I know there are problems with this analogy but I couldn't think of a better one!

In addition, a stiffer system, with a higher natural frequency, would be more unlikely to be driven at that natural frequency by a cyclic input (e.g ballscrew rotation), and less likely to go into resonance. Again the displacement is less.

When you talk about damping, this comes from 2 sources:
Inherent damping in the material (e.g wood is well damped, aluminium has some, and steel has less)
Mechanical damping at the interfaces (e.g. microscopic bolted joint movement)

You can apply external damping to panels (e.g. damping sheet) which this turns the vibration energy into heat, but this isn't really applicable for structures.

Damping only has an effect on the amplitude of the resonances. It has absolutely no effect elsewhere i.e. away from resonances. So a less sturdy machine has no way of 'damping' the vibration soley from being 'less sturdy'.

Another way to reduce vibration is to add mass. Although this lowers the natural frequency, because of Newton's Law:
force = mass x acceleration
or
acceleration = force / mass

Addition of mass, reduces acceleration during vibration.

How do I know all this stuff? Well, my day job is a Noise and Vibration Engineer and we have to measure this sort of thing, and solve problems relating to it.

Ross77
17-07-2010, 10:01 PM
Huuummmm. can of worms well and truley opened :smile: about time someone shut me up......

I'll have a wee think on re-phasing my logic and reply later, I've not a nice new x y table in the workshop that is begging to be taken apart.....:toot:

routercnc
17-07-2010, 10:11 PM
Please don't shut up, you have alot to offer and have helped alot of people out! Sometimes your keyboard is faster than your mind in your eagerness to help, that's all. I think you know this stuff really, just need time to mull it over. Have fun with the X Y table (that ebay one?)

Ross77
17-07-2010, 11:48 PM
Sometimes your keyboard is faster than your mind in your eagerness to help, that's all


:rofl:If you knew how long it takes me write a post then my mind must be slower than a snail on national go slow day :smile::smile::smile:

I certainly dont know everything (but I'm keen to fill in the gaps) I just try to look at things from a practical point of veiw as well as pages of theory and calcs.

My logic was that vibration can only be transmited through a stiff material. If you take a steel bar and tap one end then vibration is transmited to the other end, but if you took a similar bar of say rubber or wood then much less vibration will be transmitted.

Therefore only a stiff structure can vibrate, No?

Anyway its getting late, Any good books you can recommed for me brush up on this subject?

GeorgeD
19-07-2010, 06:48 AM
I would look to get some large pieces of Ally to melt as the smaller the pieces the more oxide you will produce.
Ideally you need a reducing atmosphere for ally not easy with propane as you need a lot of air to get the temperature.
I would like to try but not enough garden(boss has too many plants)

peterHi Peter

I don't think you have looked into this as much as I have?

You melt down all the small stuff together and mould decent sized? ingots out of them ready fo melting the components to be made.
The ideal Aluminium is the type thats already been casted before ie cylinder heads or any large casted component as this will be prime material.

As for Propane torch,there is lots of source out there in constructing a torch head to give maximum heat flame within the homebrew furnace,believe me its so simple.

Follow this guys series of making his own CNC,everything is home made/built/casted,such a pity he didn't have the brains to protect himself for safty when casting though.
http://www.youtube.com/watch?v=DXFDsknP0DM&feature=related

ptjw7uk
19-07-2010, 07:59 AM
I have done casting before but under lab conditions, reducing atmosphere but it still remains that evry time you melt Aluminium the surface of the pieces are scrap as the oxide will never melt.
You only have to see the prices paid for ally turnings, not worth a lot at all.
Each time you melt you produce more oxide scrap and are just wasting energy.
If you want to play then go ahead but you could end up paying more in propane than the excercise is worth!

Peter

routercnc
19-07-2010, 12:59 PM
Hi Ross,

I'll keep this brief because we're deviating a bit from George's post, and I think he's lost interest anyway!

The reason things pass vibration along their structure is due to the level of damping. Highly damped materials such as wood, don't 'ring on' or let the vibration travel very far through them because the small movement of the fibres creates heat and the wave quickly runs out of energy. Lowly damped materials, such as a steel bar, will pass this vibration through the structure without much loss. But if you add external damping to the steel, it will be just as stiff, but not vibrate so much. So stiffness and vibration can be seperate, but a highly damped structure will always vibrate much less because damping is the controlling factor for vibration propagation.

I think that the bottom line is though, getting back to George's post, is that we're all a bit worried about the use of perspex for a structural item, which is exposed to repeated vibration inputs from the cutter etc. I don't know enough about perspex specifically to comment on the stiffness/damping/cracking question, but wish George well with the idea . . .

GeorgeD
20-07-2010, 03:56 AM
Hi Guys.
Most of you are talking about vibration being a major factor here when building a CNC in certain acoustic material.

For the life of me the only area I can see vibration emmitting is on the spindle Y axis? if we were to dampen the spindle were it is clamped to the Y axis with rubber seal then vibration would be kept to a minimum.
Any type of bearing/s should glide with precision and if it rattles then its poorly made.
We could also dampen the steppers with a piece of rubber matting at the face of where they're mounted.

GeorgeD
20-07-2010, 05:18 AM
Whats the pros and cons over a fixed gantry and a moving X axis?

Since I'm going to tackle the CNC in perspex I thought it might be wise to have the gantry static rather than moving.

irving2008
20-07-2010, 09:11 AM
The up side of a fixed gantry, moving table means, for smaller work pieces, smaller drive mechanisms as (generally) you're not throwing so much weight around and a more rigid structure. The downside of a fixed gantry is a larger footprint for the same work area. A 500mm long area needs the table to move 1m, whereas with a moving gantry it traverses 500mm plus the 1/2 the gantry depth in the X-direction (often the bearing spacing) so potentially as little as 600mm.

FatFreddie
20-07-2010, 09:54 AM
For the life of me the only area I can see vibration emmitting is on the spindle Y axis? if we were to dampen the spindle were it is clamped to the Y axis with rubber seal then vibration would be kept to a minimum.

If you rubber mount the spindle, it will flex and reduce accuracy. Basically, for accuracy, you want maximum rigidity all the way from the tool tip to the work piece which means using rigid materials. Any form of damping that relies on movement will reduce accuracy, the only thing you can really do to control vibration is to increase the mass so that the energy going into the system is dissipated by moving the whole machine a very small distance. If the forces you are putting in are small, say a low power well balanced spindle cutting foam, or a laser then you do not need so much mass or stiffness.

GeorgeD
20-07-2010, 10:36 AM
Hi FatFreddie.

The rubber I was on about would be of 1mm thickness and the clamping of the spindle would then give it a tight secure grip. :-)

FatFreddie
20-07-2010, 10:43 AM
The rubber I was on about would be of 1mm thickness and the clamping of the spindle would then give it a tight secure grip. :-)

Hi George,

It all depends on the level of accuracy you want. If you can accept that amount of flex (and believe me, in some applications that would be A LOT of flex) then you should consider wood as a material. It's cheap, plentiful and as already mentioned, has good damping qualities. I've built a few things from perspex and it's not a great material either to work with or structurally.

irving2008
20-07-2010, 12:47 PM
Hi FatFreddie.

The rubber I was on about would be of 1mm thickness and the clamping of the spindle would then give it a tight secure grip. :-)

The modulus of elasticity of rubber is about 0.1GPa. A crude estimate of the flex at the tip of a typical router 50mm below the mount if this was used to mount a router in a 43mm mount 30mm deep would be about .25mm per 10N of applied cutting force... doesnt make for an accurate cut!

GeorgeD
20-07-2010, 01:09 PM
Hi guys.

This is in no way meant to be rude or a stab at your vast knowledge which is appreciated by all reading and listening to the knowledgable,but this is alight in theory most of what has been put forward but has most of it been put into actuall practice?

What I see is a lot of maths being worked out and I've alays said that the practice can sometimes flaw the theory.

Again I am not doubting or thwarting your input and its great that you take the time to help those not knowledgable about these things. :wink:

florins
20-07-2010, 01:33 PM
Hi guys....the practice can sometimes flaw the theory.

...the practice can sometimes flaw the FAULTY theory... However, the theory is well established in this field, and irving2008 (and all the others that tried to help) explained everything in layman terms (or as layman is possible). If you need to go all the way up, then go and please let us know of the result - that's why we are here.

Good luck,
Florin

Ross77
20-07-2010, 01:35 PM
Hi George

I'm going to steer clear of the vibration stuff untill I understand it a bit better. But listen to Irving and Freddie and dont use rubber at all.

I kind of agree with you about reducing the vibration but the only way I can see to reduce spindle and motor problems is to use mechanical counter shafts and balances like in a car engine or tuned mass dampers but thats all getting very silly and complicated and I will definatlly be shot for suggestting it.....:tongue:

Its sounds to me like you are at the same stage as me when I asked if it was possible to convert a pillar drill to a mill. Every one said no dont bother its too much work, but I went ahead and did it anyway ......... and yes it was a lot of work and still isnt fully sorted but I have learnt so much in the process. (and I wouldnt bother again) If its only your time then you have nothing to lose.

At least it would end the debate .......

GeorgeD
20-07-2010, 02:03 PM
Florin,why did you do that? ie quoted my text as if thats what I was actually saying.

I was just putting my thoughts across and thats all they're.

Cheers.

I'll leave the forum and go elsewhere.

Irwin thanks for all the time you gave me on getting this far,much appreciated.

florins
20-07-2010, 02:44 PM
Florin,why did you do that? ie quoted my text as if thats what I was actually saying.

I am sorry George, but it is quoted from your previous post (post #29)... maybe I misunderstood what you tried to say.

In my opinion, as a physicist, a good model replicates very well the reality, and this is the base of all human progress in the modern age. Instead of MAKING the mistakes, you can MODEL and avoid them. That's maths and physics good for.

If you can not model them yourself, you should trust the more experienced persons. Otherwise, you just waste time and money, but as Ross77 said, this can be educative.


All the best.

Florin

GeorgeD
24-07-2010, 10:36 PM
I've put the idea of using perspex to bed for the CNC build and opted for mild steel box section tubing, somewhere between 3mm and 4mm wall thickness.

Have purchased a mig welder,should arrive Monday? the thing that I now need t decide is to whether to have the gantry static or traverse?
Someone on here posted a link to a machine that was also done with box section mild steel and thats what swayed me to build it in this material,it wasso nice I wanted it. :)
Intentions of this build was to use as a router/carver and with future components changed for better quality this was the ideal material to use for heavyish work?

I won't be aquiring the box section for another week or so? but this will give me adequate time to get to know the Mig,had a go at arc welding in the past but that died a death due to me getting flashback at sometime during the usage of the arc welder,terrible thing to happen as its really very hurtful on the peepers for a day or two.

florins
24-07-2010, 11:44 PM
...opted for mild steel box section tubing, somewhere between 3mm and 4mm wall thickness.

Fantastic George, steel is a much better way to go! Good luck in your project and keep us posted!

Florin

GeorgeD
25-07-2010, 04:21 PM
I figured its much easier to cut,construct and weld the frame in mild steel than it would in working with Aluminium.

I also worked it out the price will be lower than Aluminium going by the prices on ebay offcuts that sellers are offering,but not only this it is does truly look better in mild steel.

:wink:

routercnc
25-07-2010, 07:38 PM
Nice one, glad you persisted with your plan to make a CNC machine. Steel is a better option.

You can use all that spare perspex to make:
Dust extractor plate, to attach a vacuum to
Safety screen
Side and end pieces around the cutting area to protect the bearings, drive system etc. from excessive dust
See through lid for your control box
etc.

Your choice of static or moving gantry depends on how much area you have in your workshop for the machine, because the static one will need about twice the area compared to a moving gantry type. Your choice might also depend on what you want to make. For example,
For small (say 400x400mm) very intricate work, a static gantry could be a good choice, because it can be made very stiff.
For larger work, there often isn't space for a static gantry, and the same level of detailed accuracy might not be required. A moving gantry is then often suitable.

If you go for the moving gantry, there is another choice to be made depending upon how wide it is. For wide gantrys, say >500mm, you might want to drive each leg directly rather than a central drive underneath. This is to reduce the risk of racking where the gantry skews to one side under load.

Good luck !

Colin Barron
25-09-2010, 08:54 PM
Aluminium may be an easy material to work with but in industry steel and cast iron are the prefered materials for machines. I would have thought that steel u section and box section would have been the prefered diy material because it is easily joined and heavy so has less vibration.