Epoxy granite - has anyone tested for Young's modulus?

Hi all - I am after an impartial sanity check. I was wondering if anyone has done testing for the actual stiffness achieved in your EG builds. I realise this may be rare for hobbyists, but perhaps you have made a coupon and rigged up something basic to estimate stiffness.

I ask in part because I have commissioned a test of a commercial solution - i.e., I paid for an expensive ready-to-go granite mix with correct epoxy dosing, had the coupons made by professionals, and had the coupons tested in a lab...I cannot be too specific about the manufacturer or numbers, but the tested Young's modulus was about 1/3 of that claimed.

Testing was flexural (4 point bend) - I wonder if larger numbers are in part a consequence of pure compressive testing.

In the Milli thread, peteeng is working on some testing. I believe he'll be making flexural comparisons between CSA concrete and EG, not sure though.

At least I didn't miss an obvious thread then!

I thought peteeng had focused in on high performance concrete - I hope he does do a comparison :)

Quote:

---

Originally Posted by Mud_

Hi all - I am after an impartial sanity check. I was wondering if anyone has done testing for the actual stiffness achieved in your EG builds. I realise this may be rare for hobbyists, but perhaps you have made a coupon and rigged up something basic to estimate stiffness.

I ask in part because I have commissioned a test of a commercial solution - i.e., I paid for an expensive ready-to-go granite mix with correct epoxy dosing, had the coupons made by professionals, and had the coupons tested in a lab...I cannot be too specific about the manufacturer or numbers, but the tested Young's modulus was about 1/3 of that claimed.

Testing was flexural (4 point bend) - I wonder if larger numbers are in part a consequence of pure compressive testing.

---

This is quite interesting... I'm going to send 5 test-samples to a national testing institute for strength and modulus-testing very soon... however, they have warned me that their testing method might not be applicable to EG as they are setup for testing "regular" concrete. Maybe this could be an issue with your tests aswell? I'll post the results when i have dem... I'm mostly concerned with the results relative to each other, which is why i'm trying anyways...

/Thomas

Quote:

---

Originally Posted by badhabit

This is quite interesting... I'm going to send 5 test-samples to a national testing institute for strength and modulus-testing very soon... however, they have warned me that their testing method might not be applicable to EG as they are setup for testing "regular" concrete. Maybe this could be an issue with your tests aswell? I'll post the results when i have dem... I'm mostly concerned with the results relative to each other, which is why i'm trying anyways...

/Thomas

---

I suspect they are referring to the force range and possibly the bending jig size as I have seen some very large test setups for civil stuff - my tests were done on a standard lab size Instron, not by a specialist. I'm interested to see your results :)

Many published research papers about epoxy granite.

Overall figures are not too inconsistent. Some outliers.

Treat with caution any paper that takes manufacturer figures on number and doesn't test.

Comparative tests usually more believable. Same testing rig etc. Does your testing lab have results for other materials that they can give you? E.g. how do their regular concrete numbers compare to published / manufacturers figures?

I have done the literature search, which is part of why I was surprised by my result - not sure yet if I blame the material family or the supplier.

The test house should be fine - I've used them in the past for other materials. It's a simple test.

What do EG mix manufacturer say about the results?

They should be pretty keen to sort it out

Hi Mud - a bending test for modulus can be tricky. It would be good to do an aluminium sample to the same test to check the accuracy of the lab test. What E did you get? and yes most compression tests of high solid ratio samples test high as the aggregate transfers lots of strain via contact vs through the epoxy. Tensile tests are the better path. The things that could go wrong with 4 point flexure are as follows:
1) The beam is too short and it does not follow the "long" beam or std beam flexure formula. Its then a short beam or a "shear beam" and uses a different formula
2) The calc data includes (0,0) which creates a large error in the calculation. The calculation should measure the slope of the curve not try to estimate the actual modulus. Then the slope is used to calculate modulus via the geometry. This is a common error with automated load/deflection systems
3) The depth to width ratio of the sample is incorrect and again it is then shear dominate which means the std deflection beam formula will not represent how it deflects
4) The deflection is measured via the cross bridge deflection vs its actual centre deflection. Most times this is OK but it has to be checked. ie as the cross bridge is pushing down on the two load points the beam deflects more in the middle then the top bolster does. If they attached a linear measurement device to the centre of the beam then all good
5) They need to test the beam in two directions to check the sample is not biased. "small" samples of this sort of stuff are often biased due to segregation and settling in the mould

Maybe others I can't think of at the moment.. If the test data is available happy to look at it...

cheers Peter

Granite and epoxy top published value say 40GPa average value say 30GPa so you got less then say 36/3=12GPa??

granite E=70 epoxy E=3.5 and you may have top solid volume ratio of 80% so (70x0.8x0.5) + (3.5*0.2) = 29GPa tops... (70x0.65x0.5) +(3.5*0.35)= 24GPa typical. To get 40GPa need to test in compression or use high modulus aggregate...

concrete can get higher as its E=27GPa vs epoxy at 3.5GPa and the aggregate addition is about the same so overall E can be much higher... My first test will be an ALOX/epoxy cantilever... Mould nearly done...

Quote:

---

Originally Posted by pippin88

What do EG mix manufacturer say about the results?

They should be pretty keen to sort it out

---

No answers yet, but short of them admitting an issue with the supplied materials I don't think anything constructive will result. I may name and shame if it's a lost cause.

Quote:

---

Originally Posted by peteeng

{...}

---

Appreciate the detailed comments - I don't suspect the lab has made a mistake as the unexpected values spurred a detailed sanity check, but I'll confirm if the material is so wildly disappointing or if there was a mistake.

Epoxy was on the order of 10%, and yes, E around your 12GPa guess (cannot be more specific unfortunately, but way off expectations!).

Hi Mud - did they provide a load/deflection plot? you can learn a lot about the result from that. There are three types of behavior: elastic (A) elastoplastic (B) and viscoelastic (C). You can see that if you calculate the E across the wrong data set you will get a "soft" answer. Did they test to failure?

elastic type materials use a brittle resin say elongation <4% EP use resins with higher elongation say 5-10% and VE use high elongation resins say 20%. In a flexure test the standard specs the strain range to calculate the E. I usually use the ISO14125. cheers Peter

Hi Mud - If we look at Rampfs data there is a large delta between the compressive strength and the flexural strength. If we look at the compressive as its easier then E=stress/strain so strain equals stress/E. 130/40000=130/40000*100= 0.33% so they calculate E over a very small initial strain.... since in a cnc machine stresses are very low say <5MPa then this is reasonable.... so maybe get them to provide load/def;lection graph and also recalculate on say the first 0.25% 0.5% of strain data....Peter

What is the purpose of this?

There's never going to be any reliable data for comparison like there is for metals because "epoxy granite" isn't a specified material. It's a composite with many possible recipes.

There is a large number of different epoxies with materially different properties and there is an infinite number of mix ratios with the granite.

Obviously testing the strength and stiffness of an epoxy granite mix is impractical because so much depends on the size and shape of the casting. You'd have to make a full size test to be sure of anything.

If the aim is to see how thick your epoxy granite casting needs to be for the intended purpose, perhaps you'd find it a lot easier to use someone else's machine design (that was made for a similar purpose) as a reference.

Also, a lot of epoxy granite machine bases use other materials. E.g. They can be cast around a steel frame to get the benefits of both materials.

peteeng - I'm awaiting the plots, but good info again, thanks.

Quote:

---

Originally Posted by Goemon

What is the purpose of this?

There's never going to be any reliable data for comparison like there is for metals because "epoxy granite" isn't a specified material. It's a composite with many possible recipes.

There is a large number of different epoxies with materially different properties and there is an infinite number of mix ratios with the granite.

Obviously testing the strength and stiffness of an epoxy granite mix is impractical because so much depends on the size and shape of the casting. You'd have to make a full size test to be sure of anything.

If the aim is to see how thick your epoxy granite casting needs to be for the intended purpose, perhaps you'd find it a lot easier to use someone else's machine design (that was made for a similar purpose) as a reference.

Also, a lot of epoxy granite machine bases use other materials. E.g. They can be cast around a steel frame to get the benefits of both materials.

---

Hi Goemon - if you cannot get even halfway to the specified stiffness result from a trusted EG kit supplier then that is beyond expected variability in the material family - it should be a high water mark. You absolutely can test representative strength and stiffness - just make the coupons at the correct scale, which I did. I am looking for either similar results from others to damn the material in general for flexural (or possibly tensile) members, or better results to damn something specific to my coupons.

Yes, I am concerned with custom designs which require decent, known material properties. I am aware of the design options for the use of this material. No significant investment can be made in a high performance machine using this material given the results I have got so far - no need to unpick further, other applications are less demanding and would not invest in checking the material as I have.

Hi Geoman - Machine designers need to know the stiffness of the material to design components. Sure there are variables but in my case, Muds case and any other Designer/Makers case we shall standardise on a recipe and stick to it if we know the recipes stiffness. The term EG covers a lot of territory but it will be very specific to each Maker, same as "concrete" there are many variables there as well but they do successfully manufacture and design with it. This sort of composite material is tested and characterised often. The question I and Mud is asking is; is the material 10GPa or 80GPa? and is it repeatable... testing is the only way to quantify that.. Peter

Hi Mud - another approach to figuring out the material modulus is "digital twinning". ie you make an FE model of your test and adjust the E in the model until it agrees with the results. If the material is non-linear or the test is non-linear then adjustments have to be made to fix that. But usually for low strains like we have in CNC machines the material is linear. If a suitable FE element is used that accounts for shear deflection then the geometry of the test does not matter the answer will be correct.... Peter

Obviously machine designers need to understand if their machine is going to be stiff enough for the intended purpose. I'd just have zero confidence in using any general epoxy granite data or any specific epoxy granite suppliers results.

One of the (many) things that makes resin different from metals is that stiffness is massively dependent on how it's cured and the conditions it will be used in.

In general, the optimum stated stiffness and hardness is only achieved with high temp curing that follows a specific schedule. You then add a lot more variables when you use resin as part of a composite.

Trying to figure out how epoxy granite compares to steel or aluminum in a Young's modulus table would not be massively useful for a machine designer IMO. Epoxy granite is an entirely different material requiring a different design.

For one thing, epoxy granite will not bend much before if breaks (unless you mess up the curing). If I was reproducing my Epoxy granite machine base to sell, the only data I'd have any confidence in is from stress testing my specific design.

Testing how much force could be exerted on the center of the gantry before if snaps would be far more useful.

Testing a small piece would not be that useful as it's performance is not linear. As it's mixed with stones you couldn't simply extrapolate that a 10" thick slab would be five times a stiff as 2" thick piece.

In the event that a design isn't strong and stiff enough, there are more options than simply increasing thickness with epoxy granite. It can be easily reinforced with steel tubes, for example, like they do with concrete.

Hi Goemon, sorry, I disagree - the material property information is useful to the designer. Nonlinearity is pretty negligible in useful machining ranges - most design work is linear and static. Curing epoxy isn't particularly challenging. Strength is mostly irrelevant.

I'm not advocating undersized coupons - I think you are confusing beam stiffness (which includes geometric effects) with material stiffness (which is normalised to area). Concrete reinforcement is a different game - in machining you generally try to avoid pre-stress (bolts and ballscrews aside). I'm not really looking to debate machine design philosophy, but to get the benefit of the EG you want to be careful about short-circuiting your load paths with a stiffer material.

Hi Goeman - To discuss your comment a little bit more:

Obviously machine designers need to understand if their machine is going to be stiff enough for the intended purpose. Yes very much so

I'd just have zero confidence in using any general epoxy granite data or any specific epoxy granite suppliers results. Zero is a bit harsh but it is good to be sceptical

One of the (many) things that makes resin different from metals is that stiffness is massively dependent on how it's cured and the conditions it will be used in. Actually stiffness is partially dependent on cure but strength is very dependent on cure If you select a resin that can achieve a "crisp" cure at ambient then that's all you really need. The aggregate being the dominant stiff additive does most of the work


In general, the optimum stated stiffness and hardness is only achieved with high temp curing that follows a specific schedule. You then add a lot more variables when you use resin as part of a composite. Optimum stiffness of the epoxy can be achieved at ambient if you pick one that does not need special post cure. By optimum I mean it is suitable for CNC machine parts. Having said that yes epoxy does benefit from post cure if you can do it

Re -epoxy curing. Heres the data sheet for an epoxy being used by one the forum members. It states it fully cures at 15deg C.

Trying to figure out how epoxy granite compares to steel or aluminum in a Young's modulus table would not be massively useful for a machine designer IMO. Epoxy granite is an entirely different material requiring a different design. Could you propose how to design then if it requires a different approach?

For one thing, epoxy granite will not bend much before if breaks (unless you mess up the curing). If I was reproducing my Epoxy granite machine base to sell, the only data I'd have any confidence in is from stress testing my specific design. All the companies making EG and UHPC machines regularly test their materials for strength and stiffness and use this information for machine design. Testing is done with representative samples that allow accurate stiffness results.

Testing how much force could be exerted on the center of the gantry before if snaps would be far more useful. Breaking the gantry would not give anyone any useful information about the machine except perhaps how many elephants it may hold up :).

Testing a small piece would not be that useful as it's performance is not linear. As it's mixed with stones you couldn't simply extrapolate that a 10" thick slab would be five times a stiff as 2" thick piece. Testing the right size piece is critical to get accurate modulus results small is a relative statement and we have to presume the people doing these tests know what size is required for the purpose. For instance I do calculations on very large composite yachts, say 20-50m long or more and we design on coupons 25mm wide by 120mm long which are not very big compared to the yacht...

In the event that a design isn't strong and stiff enough, there are more options than simply increasing thickness with epoxy granite. It can be easily reinforced with steel tubes, for example, like they do with concrete. Using steel in concrete is to control cracks and to improve the concretes tensile strength. Both are not needed in a CNC machine. Plus you introduce stiff loadpaths to the machine that may be counter productive

To conclude - The stiffness of any material used is needed to design a machine. The stiffness can be tested for and if the recipe is consistent and the casting is consistent then the resultant modulus is consistent. The rules or domain of static linear analysis apply and the strains in most stiff cnc parts is very small and nowhere near the failure or yield point of the material. If there is another approach to machine design, I and others would like to know about that so please share :) Peter

attached is a dat6a sheet from an epoxy being used by one of the members. It states it fully cures at 15degC but it does recommend a post cure schedule of 40C for 15h. If you need a high Tg or HDT, which machines don't unless they are operating on ovens then it needs to be high temperature cured....

Hi Mud - Any more News? I am coming to the conclusion that the high modulus numbers achieved using compression maybe a red herring on tensile and flexure modulus. I have been playing with a sample of alox/epoxy that should be really stiff but its not as stiff as aluminium yet it should be at least equal. If this is the case then I have to drop back to a known point which is thick infused fibreglass/epoxy laminates (E=30GPa tensile, compression and flexure via test) that I have years of test data on and then fill with E=30GPa concrete non shrink grout.... At least I'll know the 30GPa is a real figure. But it would be nice to have a higher mod material.... Peter

No news, but I share your suspicion.