584,860 active members*
5,239 visitors online*
Register for free
Login
IndustryArena Forum > Mechanical Engineering > Epoxy Granite > Steel structure in EG or just inserts?
Page 1 of 2 12
Results 1 to 20 of 21
  1. #1
    Join Date
    Aug 2006
    Posts
    258

    Steel structure in EG or just inserts?

    Hello,

    I'm looking to build a large'ish 3 axis horisontal mill using EG. 7,5kw BT40 spindleI, travel is X=600mm, Y=400mm, Z=650mm, rapids around 30m/min'ish.

    My intention was to first weld a structure in steel(20mm flanges throughout and 8mm braces across), get it annealed and then encase it in EG afterwards. I would encase piping to allow temperature control of the casing, both during the postcuring of the EG and grinding of guideway mounting surfaces, but also afterwards during operation...But it seems like "everyone" is just using some sort of inserts and now i'm starting to think that the inner steel structure might be a bad idea and actually make the machine perform worse than without it. I'm getting a bit nervous about the difference in temperature coefficient and that the poor damping of the steel might allow vibrations to propagate throughout the steel inside the EG....

    Any thoughts?

    /Thomas

  2. #2
    Join Date
    Aug 2006
    Posts
    258

    Re: Steel structure in EG or just inserts?

    I thought it might help to understand it better, if I posted a screenshot of the steel structure I'm talking about. Here is a screenshot of my model so far, with the steel inside the EG:
    Attachment 452960

  3. #3
    Join Date
    Mar 2006
    Posts
    42

    Re: Steel structure in EG or just inserts?

    Don't forget attachment points for your way covers and coolant surround.

  4. #4
    Join Date
    Aug 2006
    Posts
    258

    Re: Steel structure in EG or just inserts?

    Quote Originally Posted by switch2472 View Post
    Don't forget attachment points for your way covers and coolant surround.
    Thanks for your input! Appreciate any "heads-up", inputs or any other comments i can get... once the EG is poured there is no going back :-) Any thoughts on if it's a bad idea with a inner steel structure instead of just "inserts" directly on the mould?

    I'm not sure if you can make it out in the picture, but there is a bunch of holes for M12 hex-nuts. in the flanges 30mm from the sides of the EG. These are for 30mm M12 hex-nuts for attaching toolchanger, lifting points etc. And to give to mould some strength during puring. On top of that i'm going to have quite a few "blind inserts" for attaching the entire enclosure, way covers, cabletracks, proximity sensors brackets, hoseclamps etc...these will just be attached to the mould and not part of the inner steel structure...

  5. #5
    Join Date
    Jan 2018
    Posts
    9

    Re: Steel structure in EG or just inserts?

    Im currently researching the same problem. My concern is that stitch-welding the steel frame would result in tiny (>.005") gaps in the steel joints that would open and close with stress rather than transmitting strain through the joint. I haven't yet answered the question to my satisfaction. My design is currently similar to a torsion-box style welding table. The stitch welding is to avoid weld distortion, so that the frame will assemble to nearly laser-cut tolerances to avoid excessive finishing work. On the plus side, the steel structure would, at a minimum, act like rebar.

    Do you have any insight?

  6. #6
    Join Date
    Jul 2018
    Posts
    6254

    Re: Steel structure in EG or just inserts?

    Hi All - I'd use the steelwork as the the mould or shell. No use burying it plus you then have to make a mould. Take advantage of the laser cut steel to the max. I'd braze the steel vs weld to minimise distortion. Its common to make a shell and fill it even at professional level (look up Rampf company) The EG thread and my "Milli" thread is full of discussion pros cons etc... very heated in some places. Every decision is a trade off. eg why have all that light steel laser cut when you could have heavy steel laser cut for same price, weld it up and thats it? Lots of successful mills made that way. EG is quite an expensive add on exercise... If you want a max damp machine then go all the way and minimise steel maximise EG. But if you max stiffness steel then its stiffer and globally damper in steel.... have to do the numbers.

    re stitch welding. I do a lot of welded structures analysis and you can see the small difference in stitch welding vs 100% seam welding in deflections. maybe less then 1% different for a 50%/50% stitch/gap. So this issue does exist.... Investigate thermal stress relief its worth it if you can have it done. 100% welded, stress relief then final machining.... Merry Christmas Peter

  7. #7
    Join Date
    Aug 2006
    Posts
    258

    Re: Steel structure in EG or just inserts?

    Quote Originally Posted by Blockerra View Post
    My concern is that stitch-welding the steel frame would result in tiny (>.005") gaps in the steel joints that would open and close with stress rather than transmitting strain through the joint. [...] The stitch welding is to avoid weld distortion, so that the frame will assemble to nearly laser-cut tolerances to avoid excessive finishing work.
    I think it will be very difficult(impossible?) to get any reasonable level of "CNC level trueness" in any form after welding. Or at least: i won't be able to :-) My approach is rather the opposite: I've designed the structure to be machined and ground after the casting is done. This way i'm able to do some heavy welding since minor warping will have no effect on the final result. I've also wondered how much the heat-treating will affect the warping(?) I've found a somewhat local company able to both machine and grind the components after casting. So i think i'm almost ready to go on this one...

    Quote Originally Posted by peteeng View Post
    I'd use the steelwork as the the mould or shell. No use burying it plus you then have to make a mould. Take advantage of the laser cut steel to the max. I'd braze the steel vs weld to minimise distortion. Its common to make a shell and fill it even at professional level (look up Rampf company) The EG thread and my "Milli" thread is full of discussion pros cons etc... very heated in some places. Every decision is a trade off. eg why have all that light steel laser cut when you could have heavy steel laser cut for same price, weld it up and thats it? Lots of successful mills made that way. EG is quite an expensive add on exercise... If you want a max damp machine then go all the way and minimise steel maximise EG. But if you max stiffness steel then its stiffer and globally damper in steel....
    I've considered the "use the steel as mould" option. But I've previous built a large CNC router, so milling the wood for the mold would be a minor thing, plus i'd like to try to make a large EG mould as part of the "learningprocess" in this project... However, it would make the entire moulding part easier if i added a lasercut "steel shell"...and the extra "lasertime" would not cost much compared to the cost of the entire project....and there won't be any headaches with getting the mold of the part afterwards....hmm...i'll have to consider this again i think....i was hoping to order the steel in the start of january, so i'll have to think fast i guess :-) Anyways, thanks for the "spark"...

    Besides that my reasoning is that steel is much stiffer(IIRC like 5-7x) than EG, so having a strong inner steel structure makes rather good sense to me. So've i went on with that concept...

    /Thomas

  8. #8
    Join Date
    Jul 2018
    Posts
    6254

    Re: Steel structure in EG or just inserts?

    Hi Thomas - Another issue with a steel skeleton or armature embedded in a EG block is the relative loadpaths. As you say EG is considerably less stiff then the steel. If the load introduction points and reaction points are all connected via the armature the load may never transfer to the EG so the EG does not get taken advantage of. For instance if the EG were rubber (1-5GPa stiffness) it would not provide any structural stiffness (EG 25-40GPa steel 200GPa). It would damp the armature but thats an expensive and complex damping method. If your CAD system has FE it would be valuable to model the composite structure to decide what bits were working and what bits are not.... strains take the stiffest path and steel is very stiff... If EG is your interest then make a full EG part and flick the armature....Peter

  9. #9
    Join Date
    Jan 2018
    Posts
    9

    Re: Steel structure in EG or just inserts?

    Quote Originally Posted by badhabit View Post
    I think it will be very difficult(impossible?) to get any reasonable level of "CNC level trueness" in any form after welding. Or at least: i won't be able to :-) My approach is rather the opposite: I've designed the structure to be machined and ground after the casting is done. This way i'm able to do some heavy welding since minor warping will have no effect on the final result. I've also wondered how much the heat-treating will affect the warping(?) I've found a somewhat local company able to both machine and grind the components after casting. So i think i'm almost ready to go on this one...

    /Thomas
    Re-reading my post, I can see I was very unclear. The stitch welding is to avoid distortion to minimize excessive finishing work. The laser cut parts allow for assembly with minimal gaps that might open and close when the surrounding structure strains. I'm hoping that distortion will be minimal after welding and annealing. I work for a company that makes a few complex weldments to ±.030", which would be amazing if that were my result. I am trying to leave the hard work to the professionals. The last project I welded was out by more than an inch over five feet.

  10. #10
    Join Date
    Aug 2006
    Posts
    258

    Re: Steel structure in EG or just inserts?

    Quote Originally Posted by peteeng View Post
    Another issue with a steel skeleton or armature embedded in a EG block is the relative loadpaths. As you say EG is considerably less stiff then the steel. If the load introduction points and reaction points are all connected via the armature the load may never transfer to the EG so the EG does not get taken advantage of. For instance if the EG were rubber (1-5GPa stiffness) it would not provide any structural stiffness (EG 25-40GPa steel 200GPa). It would damp the armature but thats an expensive and complex damping method. If your CAD system has FE it would be valuable to model the composite structure to decide what bits were working and what bits are not.... strains take the stiffest path and steel is very stiff... If EG is your interest then make a full EG part and flick the armature....
    Agree. But isn't this the tradeoff I'll have to make? If i want a stiff machine, i'll have to use a lot of steel. If i want a machine with a lot of damping, i'll have to use a lot of EG. So making a tradeoff between the two is the academic and pratical challenge. Of course both have to balanced against praticality and cost. I was hoping that creating the internal welded steelstructure would give me "adequate" stiffness and the EG encasing/filling would give me "adequate" damping within the pratical limits. I've added holes in the steel to increase the "contact" between the steel and the EG for better energy transfer at a minimal strength-loss. I have previously work quite a lot with epoxy and glass/carbon fiber, so that's the reason i chose EG. I did consider UHPC, but pricing-wise there is not much difference.

    My original concern with this post was actually if i was going to get a worse preforming machine because of the steel structure. My concern was related to what you are describing, that the inner steel structure would "negate" the damping effect of the EG by allowing the energy a easy travelpath through the steel. I don't know if this is actually "a thing" or if i should worry about it. Traditional machines are made with large cast-iron sections, and these should suffer from somewhat the same effects then....but without it being an issue.... i guess?

    PS: I'm using Fusion360 and don't have access to FE(as far as i know)... i tried to do some stress-analysis in the beginning of my design, but i wasn't sure i was doing it right and concluded that "garbage in, garbage out" made my tests useless...and instead just beefed up the design :-)

    Quote Originally Posted by Blockerra View Post
    The stitch welding is to avoid distortion to minimize excessive finishing work. The laser cut parts allow for assembly with minimal gaps that might open and close when the surrounding structure strains. I'm hoping that distortion will be minimal after welding and annealing. I work for a company that makes a few complex weldments to ±.030", which would be amazing if that were my result. I am trying to leave the hard work to the professionals. The last project I welded was out by more than an inch over five feet.
    A friend of mine is a professional welder, so he has been looking at the design from a pratical perspective with respect to the distortion and given me the thumbs up. I'm going to do the welding at his place(better gear, him mentoring/guiding and moral support during the many hours of welding). So i hope he'll be able to guide me towards lowest possible warping...

    Thanks for all you input!

    /Thomas

  11. #11
    Join Date
    Nov 2013
    Posts
    4282

    Re: Steel structure in EG or just inserts?

    Hi,
    rigidity trumps damping every time.

    If a machine lacks rigidity the machine will deflect rather than force the tool to engage the material to be cut. Toolpath accuracy=zip, surface finish=zip.....
    who cares about damping in face of that?

    Bang for your buck, a well designed welded steel structure is the cheapest most rigid construction there is.

    I had grey iron beds cast for my new build mill. What with the casting and then sending them out for machining is not cheap. The frame I had hoped to cast is SG iron,
    but I've all but run out of budget. I suspect that I will perforce make a welded steel frame to hold the cast iron beds.

    EG is not cheap either, not just the materials but the molding and inserts are very time consuming. Nor is EG, except in really thick sections particularly rigid, ie you have to invest
    big dollars to match the rigidity of either a steel structure or cast iron.

    Craig

  12. #12
    Join Date
    Jul 2018
    Posts
    6254

    Re: Steel structure in EG or just inserts?

    Hi Thomas - Agree with Joe above. So if you are to laser cut everything and really want to overcast in EG, then laser cut and bend everything and bolt it together. Zero welds. Then overcast. The bolting allows adjustment, bolted joints are damp and then it all gets frozen in the cast. I have made 4 routers all laser cut with flanges and bolted together. System works very well. Make armature as light as possible. The bulk size of EG will be the stiffness. Revisit FE in F360 its worth the effort for the insight it gives. In your case model the steel and EG separate, load them the same to figure out the relative stiffness of the two. Then if your up for it you have to model the EG, boolean the armature so it becomes an exact hollow then add the armature back in and load it up. F360 may even do a modal analysis so you can pick up vibration modes (which will be zero in the casting being such a chunk, at least modes that are not of interest for a machine like this) There are heaps of successful 100% steel machines in the forum so rigidity trumps dampness and machine cannot be too stiff.

    If your keen on EG you have to do some costing and tests because if you choose sand and epoxy you will not get past 25GPa stiffness and then you have to use heaps. Your in Europe so look up Rampf known product with known modulus very helpful....Peter

  13. #13
    Join Date
    Jul 2018
    Posts
    6254

    Re: Steel structure in EG or just inserts?

    Hi Thomas - To discuss damping. Damping is the diminishing of vibration via internal friction. We know that EG is about 20x damper then steel. A machine is a spring cantilever so when it deflects and releases it springs back. Inertia then carries it the other way then it stops and springs back. This cycle (a tuning fork in effect) may take 20 or so cycles in steel to stop as a free vibration. In EG it would take a couple of cycles. This damping is due to internal friction turning micromotions into heat (same as rubbing your hands together). So in terms of a machine this vibration can result in chatter as the tool engages and springs back from the surface its trying to cut vs being rigid enough to dive straight into the metal. This initial deflection has to be much smaller then the chip load distance. Look up chip loads for steel and aluminium and you will see they maybe 0.5mm or 1mm... If the chip load is say 1mm then the machine cannot deflect more then that other wise the tool can't push into the material. So if you know the chip thickness and the cutting force this will give you some idea of the stiffness required. Machine static stiffness of commercial VMC starts at about 70N/um (1um = 0.001mm) I'm aiming at 10-15N/um for my bench mill. Other posters have measured their machines at about 10N/um and light bench mills at 1-2N/um are usually unsatisfactory with medium to heavy cuts. Special heavy machines are in the 300-400N/um range. So figure out your machine stiffness it will tell you where your at, then consider damping... If its stiff enough to cut then it does not need to be damp.

    With a light finishing cut the chip thickness is even less and to get good surface finish the machine has to be very stiff even though the loads are smaller to allow the tool to start the cut vs bouncing along the surface. Peter

  14. #14
    Join Date
    Nov 2013
    Posts
    4282

    Re: Steel structure in EG or just inserts?

    Hi,
    when I was thinking about my new build mill I considered EG, cast iron and steel.

    Of the three a steel structure, either welded or bolted was the cheapest solution for the rigidity I wanted....but damping?

    To achieve the same rigidity in EG was going to take really large volumes of EG and I came to the conclusion that it was about double
    to triple the cost of a steel structure.

    I decided therefore to cast iron beds....which I duly did. The pattern making and pouring are not cheap, it worked out to about $10NZD/kg,
    but the result is good. Then I had them machined at a company that specialize in large format machining almost exclusively on late model Okuma machines.
    This was eye-wateringly expensive, so much so that I now consider the cost/value ratio of cast iron to be questionable.

    The results of casting your own beds is very VERY good indeed, dimensional, rigidity and damping. What I did though was underestimate the cost,
    and by quite a margin. By my estimate cast iron costs as much and more than EG, that is to say triple what a steel structure would cost.
    I have no doubt that if you were manufacturing in volume then casting and subsequent machining would come down markedly in price but as a
    one off the cost is very high despite the great results.

    I have not done anything in the way of a scientific comparison, this is however my conclusion:
    1) Steel structures are cheap and rigid
    2) EG is twice to three times the price by virtue of the large volume of EG required to get the same rigidity
    3) Cast iron has great rigidity, damping and dimensional stability but as a one-off is the most expensive of the three
    techniques, mitigated only by manufacturing volume.

    As I posted earlier I've all but exhausted my budget....so I have three cast iron axis beds, beautifully and accurately machined, three 'you beauty' C5 double
    nut ground ballscrews with my own mounts and yokes, three surface ground saddle plates, three sets of 'new old stock' heavy duty THK linear rails and cars,
    three 750W Delta B2 series servos (one braked), home and limit switches with my own mounts and actuating cams, nearly all assembled, shimmed and ready to rock.

    I was lucky enough to find that a company whom supplies me with electronic parts had bought second hand a Yamaha pick-and-place machine for parts.
    Once it was stripped down the extremely heavy duty table left over (1.3m x 1.1m on castors with adjustable feet weighing 300kg) was going spare and they gave it to me.
    Most unusual but welcome Christmas present ever!

    I'm going to use the same control system (Mach4/ESS/MB02x2), the same (two) spindles and the coolant pump/reservoir as are on my mini-mill currently.

    All that I lack is the frame to bolt the beds to.

    I have had three design ideas:
    1) A single SG casting, approximately L shaped 1.2m x 0.9m and 250mm wide. The section is such that it would weigh approx 275kg.
    Pattern making and mold making would be expensive for this design, but would not require (nearly) any machining when done.
    2) Two piece SG casting, approx 350kg. The two pieces would bolt together at right angles requiring that the mating surfaces be machined
    flat, square , drilled and tapped, but present much less challenge pattern making and mold making.
    3) Steel structure, two L shaped steel plates 40mm thick held 250mm adjacent one another by a combination of welded and bolted steel
    plates/gussets.

    Of the three I suspect that the steel structure will win out in price, I will do some drawings and submit for pricing over the coming weeks.
    It may be also that I go with a steel structure now but plan for an improved cast frame in the future.

    The main reason my budget has suddenly tightened is that I'm buying into the business for which I work. This is a great oppurtunity for me
    but has absorbed all my extra capital. None the less the business will be the primary beneficiary of my new machine, my existing mini-mill
    is doing sterling service for the company as it is, and a new, larger and more rigid machine will do better again. Maybe the business will
    stand the cost of cast SG frame over the next year or so.

    Craig

  15. #15
    Join Date
    Aug 2006
    Posts
    258

    Re: Steel structure in EG or just inserts?

    Quote Originally Posted by peteeng View Post
    If your keen on EG you have to do some costing and tests because if you choose sand and epoxy you will not get past 25GPa stiffness and then you have to use heaps. Your in Europe so look up Rampf known product with known modulus very helpful....
    I've done quite a few tests with different "fuller curve matched" batches of different sorts of locally available stone/gravel/sand. I also bought different grainsizes of aluminumoxide and made fuller curve matched tests with those(pricy!). This is kinda interesting because AlO2 has a youngs modulus above steel, 215-400 usually, so even combined with epoxy it would be very stiff if the mixing is done correctly, but the downside is very high cost. Finally i bought Silimix 282, which has impressed me a lot. It becomes much harder and much more ridig then my own mixes of stone, for some reason. So either my fuller curve mixes are all wrong or they are doing something differently, most likely "better" grainsizes. I've seen a few make a "just sand and epoxy" solution for their machines/projects, so i also tried that and the results were VERY poor compared to the properly made mixed grainsizes. The Silimix282 was on a whole other level compared to the sand samples. Sand+Epoxy is a no-go for me.

    I also contacted an epoxy company(R&G in Germany) and they recommended an epoxy for this project which i've tested, also with good results...only downside is the post-curing which would be 70c for at least 15hours... i'll do that be putting some piping in the casting and send hot water through... but the resulting casting had very little trapped air compared to the infusion resin i've normally used, which surpriced me quite a lot...

    I've pretty much decided on the Silimix 282 and the R&G epoxy since this gave me the best results in my tests. And it's not insanely expensive as a aluminium oxide mix would be. The Silimix also comes in 25kg bags, which would be easy to handle with a concrete mixer and a fixed amount of mixed epoxy per batch....it's practical, just a lot of bags and a lot of batches :-)

  16. #16
    Join Date
    Aug 2006
    Posts
    258

    Re: Steel structure in EG or just inserts?

    Quote Originally Posted by joeavaerage View Post
    when I was thinking about my new build mill I considered EG, cast iron and steel.

    Of the three a steel structure, either welded or bolted was the cheapest solution for the rigidity I wanted....but damping?

    To achieve the same rigidity in EG was going to take really large volumes of EG and I came to the conclusion that it was about double
    to triple the cost of a steel structure.
    Did you consider the combination like what i'm aiming for? Strong inner structure for stiffness encased in epoxy for mass and damping. If so, what deterred you from going down that route?

    /Thomas

  17. #17
    Join Date
    Jul 2018
    Posts
    6254

    Re: Steel structure in EG or just inserts?

    Hi Thomas - How are testing for stiffness? Peter

  18. #18
    Join Date
    Aug 2006
    Posts
    258

    Re: Steel structure in EG or just inserts?

    Quote Originally Posted by peteeng View Post
    How are testing for stiffness?
    umm, i had a kinda "unscientific", but pragmatic, approach to that. Simply casting a sampling in a specific mould and then seeing how much the sample would deflect under load. I could have made a spreadsheet and measured the deflection vs. load and calculated the youngs modulus, but i just compared them to eachother by hand...

    I also cast a sample specificly for breaking to see how and where it cracks when i broke it with a hammer....and to see how much air was inside. This was where the silimix especially shined...

    One major thing for me during my testing was also how the mix was to handle when wet. Would it flow? How would it handle in a mould? how much compacting/vibrating is needed etc...10-12% epoxy by weight seems like a good spot for most of my tests. 10% requires more work and trap more air. 12% can almost flow by itself into the mould... again i wasn't going for maximum strength, but more a good filling of a mould since i'm aiming for damping... my current plan was to do the calculations for mixing a 12% and a 10% batch. Then start with 12% batches and then maybe switch to 10% batches if i feel like it is too wet during casting...or maybe alternate between the two... i dont know yet

    /Thomas

  19. #19

    Re: Steel structure in EG or just inserts?

    badhabit, you are absolutely right! no need to connect inserts. this connection will ruin the machine, all vibrations will be transmitted from insert to insert. But if you don't take care of the adhesion of steel to epoxy granite, then that can be bad too.

  20. #20
    Join Date
    Nov 2020
    Posts
    361

    Re: Steel structure in EG or just inserts?

    Quote Originally Posted by badhabit View Post
    I thought it might help to understand it better, if I posted a screenshot of the steel structure I'm talking about. Here is a screenshot of my model so far, with the steel inside the EG:
    Attachment 452960
    Absolutely beautiful piece of work, its like art but better because its functional. Unfortunately based on research, not how they(big companies) use EG. Like Pete said, steel frame filled with EG + inserts is the general consensus.

Page 1 of 2 12

Similar Threads

  1. using Polypropylene for cnc structure?
    By garyocampo in forum Mechanical Calculations/Engineering Design
    Replies: 4
    Last Post: 04-14-2015, 06:27 PM
  2. Specially Designed Small Steel Structure CNC Router 1325
    By Baiansy in forum Roctech CNC Routers
    Replies: 0
    Last Post: 03-28-2015, 11:58 AM
  3. Structure without welding
    By metalltwist in forum Open Source CNC Machine Designs
    Replies: 2
    Last Post: 12-16-2014, 05:18 PM
  4. Moose structure
    By cncadmin in forum EnRoute
    Replies: 0
    Last Post: 05-06-2012, 04:00 AM
  5. Hi speed inserts for stainless steel
    By gearsoup in forum CNC Tooling
    Replies: 1
    Last Post: 12-29-2007, 04:07 PM

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •