Printable View
Hi CBF - I haven't made a stand like that before so can't give direct experience. But "thin" free edges and large flat plates do vibrate. Someone else will have to guide you on this one! If the unit is used for the coolant then the coolant is acting as a damper... Peter
Pete,
No worries, I appreciate all the input you have provided so far. I am not opposed to having the coolant tank be attached to the stand, as long as its removeable. Like if the tank was on a track that allowed it to be slid out like a drawer. Would this aid in the dampening?
If the track cuts down the panel size that could work... unfortunately a lot of this stuff is trial and error. Build what you can then fix the issues as they arise :) peter
No, a removeable tank or chip tray would not help with dampening, it can cause more resonance, which would only affect the tank or try and not your machine table, you won't have anything much from your table down, that will affect machining, because you are using that big Granite Table, depending how you make the side extensions ( rail supports) they will have some resonance which will carry through the Gantry Beam, the Beam will be like a drum so resonance from the spindle Z axis will extend to the Gantry Beam also, the Table is the only damp part of your machine same as with the Datron machine that uses the Granite TableQuote:
Originally Posted by ChipsBFlyin
That is why you see a lot of the demos with the Datron machines have the work is fastened to the Granite Table, you can see in this video the resonance in the cut at high speed, with adding a finish pass it will be smooth finish
https://www.youtube.com/watch?v=uswnasjwQsM
Mactec,
Those roughing roughing passes looked fine to me, when I do 4 and 5 axis work, I usually have a long stick out on both work and tool, those sets ups tend to be less rigid, but I just bring it to size/finish with the finish pass.
Anyways, thank you for your input, but I maybe going back to the drawing board after getting some quotes today, I may not be making the Y axis arms out of steel. Anyone have any input of what's the average cost of epoxy granite per cubic foot? Looks to me like the most expensive item of epoxy granite is the epoxy.
Hi,
the problem with epoxy granite is that its just not that stiff.
Youngs modulus of epoxy granite, depending on the mix and other items is in the range 35 to 70 MPa. Cast iron around 110 MPa, and steel 200 plus MPa.
So to match the stiffness of a 50mm x 50mm (solid) piece of mild steel would require approx 100mm x 100mm in epoxy granite to achieve the same result.
(The stiffness of an object varies as the square of its dimension...as a first approximation)
Considering the volume increase then steel start looking dirt-cheap by comparison.
It is true that machine manufacturers are increasingly using composites in their machine frames.....but try to find a cheap one!!! The cheap end of the market is still dominated by cast iron and/or
steel.....and the reason is simple, these materials result in the stiffest machine per $1000 invested.
Craig
Hi CBF - Now that's a hard turn leftish? I suppose not looking t #1 entry
Epoxy granite is quite expensive and getting more expensive (petrochemical product) plus its modulus is really an unknown. Many suppliers quote compressive modulus only which is skewed. I'd expect not to achieve better then 30GPa if you use sand. Are you in north amercia? Engineering grout (CSA concrete or UHPC) is the cheapest material via volume but it to is not very stiff around 30-40GPa but you can make up the delta via bulk. Grout comes in non shrink variety's. I think it out performs DIY epoxy grout. It pours into the mould easily and the tests I've done so far are encouraging. Steel or aluminium inserts can be used for post machining. Or you can make recesses for bonding in inserts then machine them. You can get self levelling grout as well but I have not tested it yet for how "level" it sets. Something like attached is good if your in NA. Elsewhere seek the equivalent. I'm about to make some carbon fibre saddles that will be interesting. Via test its about E=60GPa Peter
My local supplier of CSA will guarantee no change in shape for 100 years so I think its good to use. I have epoxy blocks here over 10 years old with no detectable change so its a stable product as well. been using epoxy for over 30 years for composite structural parts...
https://www.ctscement.com/product/ul...?c=PRODUCTS&t=
My Milli thread covers a lot of territory in composite & mineral cast construction. Casting parts has lots of advantages, that's where I'm heading vs metals
Craig,
Thanks for your input. I am trying to consider all of my options. The cost of the Y axis rails mounts, including material, SR, machining/grinding is at least 2k+. Then I still have the dilemma of mounting them to the surface plate. I checked the surface plate and the sides are not square to the top to themselves or anything else. They aren't even flat, so my best solution for that so far would be to add jack screws to the Y axis mounts and use them to offset the steel mounts from the surface plate, then proceed to snug down the mounting bolts. I would then fill the gap between the surface plate and the y axis mounts with epoxy. I am questioning how this mounting method would work, so that's when I started looking into epoxy granite.
Pete,
I am in Brooklyn, NY.
I have no idea what material I would use for a casting yet, I have never heard of the grout option, but would like to hear about it. Is it prone to cracking? I do like the idea of UHPC, but I thought the concern with concrete is that it never fully cures and tends to move a bit? I don't know enough about these materials, I still have a lot of research to do. I did stubble upon a company called Basetek and they make castings, but based on the picture I see on their website, they don't look cheap.
I've seen some people mix in fiberglass and some mentioned carbon fiber into the epoxy granite. Any benefit in that?
Any recipes or commercially available products that have been tested would be greatly appreciated.
My process for making my casting would be:
1 - Build a mold around the surface plate.
2 - Hold the inserts in a machined jig that I can adjust to set as parallel to the surface plate as possible.
3 - pour the mold
4 - self level epoxy the Y axis rails mounts
5 - hand scrape the self leveling epoxy of the Y axis mounts to the surface plate.
Anyways this is what I have in mind so far.:
Hi CBF - I think you will be disappointed with epoxy foundations. Epoxy is about 3GPa stiffness so its effectively rubber. I know its done but there are commentaries out there that its not a good solution. You need to do it about 1/4" thick to achieve "self levelling" at this thickness its way to soft. Thinner and it crawls and gets surface waves in it due to differential curing and thermal effects... Setting parts with epoxy is a different matter.
CSA grout is about 30-40GPa stiffness and there are many European machines made with similar materials. NASA has a concrete CNC project on at the moment. Engineering grout (CSA concrete vs portand concrete or cement) does not crack or move once cured, different animal to portland cement. My local supplier has given me lots of test info on their CSA and its very stable. I imagine the CTS rapid-set stuff is similiar. Does your machine have to be that big? having the surface plate means you can cast precision surfaces off of it. Wax it up, box up and the part will be as good as the surface plate. Just means you can't make the precision parts bigger then the plate. You can make CSA parts and lap them to the required finish. Buy a bag of grout and have a play. Its not costly being a building material. The supplier will be able to give you its modulus etc. Leave the surface plate as a casting surface and cast your new machine off it.... Concrete has same coefficient of expansion as steel so steel is perfectly compatible for inserts or associated parts.
Here's the data sheet for the material I have been playing with. Its spec E=34GP. I have test data that shows it 40GPa typical... Its tough to get EG to 40GPa using sand/granite.... casting is controllable and a very good way to go IMHO. The only material I can get to 80GPa is vacuum cast carbon fibre. I played with heaps of materials over the last 20 years.... Commercial CF is too expensive for general machines that I sell but I collect CF offcuts from two customers so I have enough to build some parts from....Peter
read the Milli post I trial steel fibres, ceramics and various bits and bobs thru that.... You have to add a lot of fibres to improve stiffness. This presents issues, I feel to KISS use something out of a box or bag that works. Trying to invent a material is a hard thing to do.
Hi,
the one super flat and true surface you have is the top surface of the surface plate......so why try to use the sides of the plate?. I can understand not wanting to place anything on the surface and
thereby reduce the work area....but as you have discovered the costs of doing it by using the sides is likely insurmountable.
If it were me I would mount two steel 100 x 50 x 9 or 100 x 100 x 9 down either side and mount the rails on the top of those. If you took the pieces of RHS to a machine shop
it might cost a box of beer to have them put in in the surface grinder to make the top and bottom of the section flat and parallel to each other.
Go to your local scrap yard and ask if they have any cast iron elevator weights. I paid $90 NZD ($65USD) each for three cast iron weights 1000 x 200 x 125.
You'd be amazed at what you can mill out of a lump of cast iron like that. You might for instance be able to make the uprights for the gantry.
Craig
Hi CBF - why not cast your machine bench? You will need a bench either way so that will give you experience with the stuff. Peter
Craig,
I agree with you, it makes more sense to mount it on top of the plate, but I would drastically reduce the working area on both X and Y, to the point that I would no longer want to build the machine because I have access to VMCs that have those work envelopes. If I do go that direction, mounting on top of the plate, I would look for a bigger plate before attempting to go this route, but I may end up at the same expense. I would then still need to make the columns and have them machined, so added cost there too. I would also need to find a surface plate that is at least 6'X5', kinda of odd ball size.
Scrap yards around here don't want to sell you any material. I've tried to buy material and machines from them in the past and it was hopeless. The only thing they will sell you is rolls of copper wire.
Pete,
I began reading your Milli thread, still have a long way to go. I'd be game to cast my machine stand first, but I would like to optimize the design so I use the least amount material necessary and not just cast sold block. I have to figure out my gantry machine dimensions first, before I start working on the base. As I understand, its a big no no to use rebar or any other steel structure inside of the machine casting, due to thermal growth of different materials? I'm guessing the base for the machine may benefit from some rebar?
I am curious about the "no change in shape", I wonder what no change in shape within an architectural structure means, maybe to them not changing is +/-1/6"? I'm just speculating here with nothing to back it up, I will continue to read about it. Thanks for pointing out this material, wouldn't known about it.
Don't forget the Ballscrew and motor mounting the part you have is easy mounting Ballscrews is just as important as RailsQuote:
Originally Posted by ChipsBFlyin
Mactec,
I have not forgotten about then, I just can't get past the main frame design, once I am comfortable with that, then I can start adding all the details.
I did stubble upon the Abel CNC machines while I was reading Pete's Milli thread. I like the designs and how the rapids they have.
Hi CBF - All materials change shape. Granite has had a million years to let all the thermodynamic and chemical events to run down. But it still has thermal changes like other materials. Metals change as well, they have internal thermomechanical issues that make them change shape. This is why they age cast iron out in the lot or heat treat it to accelerate the processes. Looking at the HE80AG data sheet after 730 days the compressive strength is 160Mpa and its dimensional change at 56 days is 550ustrain shrinkage. I have some 1000 day data here but can't find it but its around the 500ustrain from memory... Just like metals the grout can be post cured at higher then ambient temp to improve properties and stabilise faster. But I think thats overthinking it. ssso 500/1e6*1000mm=0.5mm per meter in 60 days. So if you cast the part leave it 2 months it will then be "stable". If you post cure at 10degs above ambienty it will be 30days at 20deg above ambient it will be 15 days to cure fully. 30C and its 7 days (say 60degsC)
If you use inventor? or Fusion It has a form finding module, very interesting to use.... That will optimise your shape.. Peter
Hi CBF - In regards to steel in concrete they have the same CoE. What people do is they connect the main elements with reo and this is poor as then the load paths are thru the steel, especially the vibrations. So if you reinforce do not connect the mesh to any of the hard points.
https://www.rampf-group.com/en/produ...neral-casting/ look thru this companies stuff to get an idea of their mineral casting tech. Peter
What I suggest is you start with a smaller machine. May not be the size you want but where your headed is a very large, very heavy machine that will be difficult to deal with. Start small so the development is fast and easy. Then you can move onto the big machine once your thru some of the learning curves. You still have heaps of hoops to jump thru size independent and its much easier to jump thru them or side step them on a small machine. Battling size creep is always an issue we want big machines but in reality we may use the entire bed once a year so maybe smaller is better for a learner.... unless you have definite job to do thats big. You say you have access to VMC's so think about billet machining of large scale aluminium parts. Al is cheap and if you go monolithic its like a casting plus everything is done on a machine so is accurate? People tend to think in terms of lots of small parts but a big part on a vmc can be cheaper then lots of small parts. Peter
Pete,
Thank you for your input. After giving it some thought, I may not go through with this project. I am not trying to be negative, but I have no intention in making a machine with a similar or smaller work envelope than the one I already have. And If the chance of success is low on a big machine, then I don't want to waste my time. I have made small machines with granite plates and aluminum blocks. Fun builds, I learned a couple things, but I ended up with thousands invested in a tiny envelope that I couldn't really use for anything useful.
I haven't given up, but I am weighing the pros and cons. Luckily I haven't really invested much into the project.
Anyways, I reached out to Rapid Set about their Ultraflow precision grout product, and it looks like its a special order item. With a full pallet minimum order. So I looked around and I found this:
QUIKRETE® Non-Shrink Precision Grout (No. 1585-00)
https://www.quikrete.com/productline...isiongrout.asp
I am not sure if this is a similar product or if would be useful for my application. As far as casting and waiting, I patient, I am happy to wait 2 months for it to fully cure. I'll keep researching in the mean time, worst case I'll post my surface plate for sale.
Hi CBF - I have many many machines that have not made it off the hard drive (probably 100's). I also have 2 small machines here that I'm wondering what to do with them (I think one will become a 3d printer). So I understand the thought bubbles. Keep the project as a paper project and push it as far as you can. The quickcrte is a combo portland/csa cement see attched. Talk to the quickcrete people they may have a 100% CSA grout. Ask rapidset about a distributor or user nearby you someone will have a bag around to play with. I turned up at Blueys warehouse and the rep gave me a bag of HE80.
The quickcrete has 2mm shrinkage per meter (2%) so a bit more than the HE80 here at 0.5mm/meter. Keep at it. We all are chasing some sort of holy grail... you may find it first. :) Peter
by the way the non-shrink data does not say if it shrinks or grows. maybe it grows.... as they say its non-shrink
Pete,
Have a look at how they pour this machines base. I'm not saying that this is the route I am going to take, these are just commercially available small CNCs
https://www.langmuirsystems.com/mr1/base
Hi CBF - Yes I've seen that. They do not spec the "concrete". But you can make light steel structures and fill with concrete. Valid approach saves making "moulds"... better then EG as the CoE is closer and the "concrete" is cheaper. Pick your poison and chase that grail. My grail is a cold castable machinable material that has E=70GPa. Only material so far to achieve this is carbon fibre E=70GPa plus, but CF and epoxy are many $$$. My current view is to vacuum infuse fibreglass thick skinned parts say 12-20mm thick (E=30GPa) so I can machine the FG then back fill with CSA grout E=30-40GPa. The compromise is lower E. Machine FG with diamond tooling...Peter
https://www.facebook.com/ScootCNC/vi...89695705018524
Re: making steel boxes and filling with concrete. Just have to say you have to be careful with this approach as the concept may just be adding mass which helps. But if the structural loadpaths are tied together via steel then you have the situation where the loads travel thru the steel not the fill. Steel is 200GPa and concrete is 30GPa so the load prefers to go down steel not concrete if in parallel as its stiffer. Plus if the steel and concrete are not "bonded" they become separate structures so then the loads definitely stay in the steel so the concrete is just added mass. So if the aim is just to add mass you may as well use more steel (steel 7800kg/m3 and concrete 2100kg/m3) as then you are adding more stiffness as well. Part of the MR-1 machine issue is freightage so its easier for them to have the Maker add concrete at point of build vs at point of manufacture. Nothing is simple in machine building it seems...
off to make some Frankie parts in formply low tech!!