[Texasbowhunter]

I have been lurking around here for a while and asked some questions early on with a build I was considering building...During this conversation there were things mentioned I had not considered and have kind of put that design on the side burner for the time being...So Fast Forward now and I saw another post of a member here that had built a Concrete CNC-https://www.cnczone.com/forums/diy-cnc-router-table-machines/434442-diy-concrete-cnc.html...
Personally, this made a lot of sense, but I did not understand at the time why he did not go epoxy/granite build until later in which I learned the cost of the epoxy in relation to just how much would be needed...So, the idea of the concrete makes so much more sense...
So, while I'm trying to absorb as much as I can on this type of design and what's involved before I decide to go down this street, I have some very early on concerns I'm unable to satisfactorily overcome but I think I'm on the right path but wanted to get some other views...

Being a hobbyist and nothing in my shop/garage is flat in all directions. I do have a heavy steel shop table in which to build the base upon, however it is not dead nut flat in all directions... So how do I establish a flat foundation in which to build the frame upon and then ensure after the pour has cured that the beams in which the rails will sit upon are as flush as one can make under the circumstances...
So, what I have considered is the table is plenty big enough for the base dimensions of the frame of the CNC...
To start with get the top of the table as level as possible and if that means shimming under the various points then so be it but get it close with a level...
Then, what if I was to lay a piece of Wood/Formica down on top of the work bench and add a 1-1/2" frame to the edges of the Wood/Formica and then pour floor leveling concrete inside the area and allowed it to self-level to give me the best possible all points level foundation in which to build upon?
Again, Just tossing this 1st hurdle around of thought since I don't have access to a dead nut foundation in which is needed to lay the frame upon and then pour and cure...
Thanks
Paul

[peteeng]

Hi Paul,
You can use self levelling grout to level the top of the base. Make the mould, fill it to say 5mm short of full, let cure then self level it out. Your best to use engineering grout for the parts not "concrete" portland cement concrete can and will crack due to its shrinkage. Use a class C grout such as sikagrout ultra or sikagrout 3350 (this is a high modulus grout) Peter

You could build a "level" table this way and then cast from that. The linked build built the base as a U but it can also be built as a base and two walls. My thoughts are to make the walls separate at the moment as I can cast them side by side (on the level base) , then use leveller which means both will be the same height, then bolt or epoxy on the base in place...

https://youtu.be/4lE81uPfA64

[ardenum2]

If you're using cheap concrete than yes concrete is going to be cheaper but if you're going for uhpc with all the magic inside, it is not going to be cheaper than EG unless epoxy is stupid expensive in your country for whatever reason. Like Pete mentioned if you wanna use portland, than it's not going to work.

another thing to note is aluminium will rust when in contact with concrete so you'd need to either cover it with epoxy at the joint or use steel, preferably stainless.

[peteeng]

Hi Ard - Have you done the numbers to show the cost difference? If you use free sand then yes the EG could be cheaper, but if you buy the aggregate it maybe not? Peter

sikagrout ultra is $34 per 20kg at local hardware and makes 23kg of solid. density=2200kg/m3 so makes 0.0104m3 ie its $3269AUD per m3. If we assume that we get sand free and that we add 10% by weight epoxy to the mix and I pay $22AUD per kg or $22AUD per litre (as the epoxy is 1000kg/m3) then bulk density of sand is about 1700kg/m3 so add 170kg of resin - say $3740AUD of resin in the 1 m3 mix...I think the grout is ahead...

[ardenum2]

I did for my region, https://www.moertelshop.eu/ is the only reliable source of uhpc around here, for anyone who wants to follow durcretes recipe. I compared it then to the cheapest but good quality low viscosity resin from https://shop.hp-textiles.com/shop/en/Epoxy-Resin/ the price was more or less the same.

tegno cement (portland/csa mix) + all the required uhpc additives like shrinkage reducers, pce etc. more or less equal the price of the epoxy, filler in both cases silimix.

[ardenum2]

Hi Ard - Have you done the numbers to show the cost difference? If you use free sand then yes the EG could be cheaper, but if you buy the aggregate it maybe not? Peter

sikagrout ultra is $34 per 20kg at local hardware and makes 23kg of solid. density=2200kg/m3 so makes 0.0104m3 ie its $3269AUD per m3. If we assume that we get sand free and that we add 10% by weight epoxy to the mix and I pay $22AUD per kg or $22AUD per litre (as the epoxy is 1000kg/m3) then bulk density of sand is about 1700kg/m3 so add 170kg of resin - say $3740AUD of resin in the 1 m3 mix...I think the grout is ahead...

I did for my region, https://www.moertelshop.eu/ is the only reliable source of uhpc around here, for anyone who wants to follow durcretes recipe. I compared it then to the cheapest but good quality low viscosity resin from https://shop.hp-textiles.com/shop/en/Epoxy-Resin/ the price was more or less the same. This resin is about 30% cheaper than the one badhabit used.

tegno cement (portland/csa mix) + all the required uhpc additives like shrinkage reducers, pce etc. more or less equal the price of the epoxy, filler in both cases silimix.

EG has no microcracking issue so I'd go with EG for longevity, but using cheap group for a first time machining sounds fine, after all, the expensive stuff like motion and servo's can be reused for the next machine. Me personally, I want to avoid the handwork associated with making 2nd, 3rd, 4th machine and get it right the first time, my health isn't what it used to be I don't think I could do it more than once.

[ardenum2]

I just love this forums server synchronization issues, double posts, delayed copies...magic

[peteeng]

I suppose with a 500,000 people on the server its expected. Peter

[Texasbowhunter]

After doing some more reading and watching the various videos your both right...Concrete is not the right option for this, but the Grout is certainly a viable option, and I should have mentioned this as well in what I was considering...
I do like the 2 options here and don't believe I would be disappointed with either option as to whether it be the EG or Grout
I personally like the idea of the fluidity of the grout to ensure a nice smooth surface...Downside is can the color be changed?... Not that the Lt. Gray would be a deal breaker...
I do like the epoxy finish though giving that appearance of a deep glass appeal but the best I can tell is the cost a gallon on the part mix...Brands to go thru in deciphering which is the best for the purpose...
If I figured up what I would need correctly, I have approx. 80 square feet to fill...
So, at @ $25 a bag for the grout and at 50lbs per bag I figured up @22 bags just for the base/walls of the CNC...so the total cost of grout would be @ $550 which doesn't seem unreasonable...
However, at this time I'm unable to determine the @ cost of the EG/ES...and add the cost of the sand to the mix...Still trying to understand this as far as 1 gallon = square feet? then when I add the percentage of sand how much will the volume increase? then how much will I need? So, more study and research continues?
Paul

[ardenum2]

you do the recipe calculations by %wt, which means eg. you have a part that's 100kg's, in case of EG, depending on your filler material, epoxy content comes between 10-15%, any less and its to dry, any more and you loose on stiffness because there's too much epoxy. So if we assume the middle ground of 12.5% that would be 12.5kg of epoxy for 87.5kg of filler(sand+gravel). Epoxies have different viscosity values and you want one with low viscosity, which means it flows smoothly and isn't too thick.

I'd say if you can get something like a premaid filler mix it would be ideal, something like this https://www.moertelshop.eu/buy-SILIMIX-282-cheaply_2. Otherwise you'd have to make your own mix of sand/gravel. There's plenty of research papers around there to guide you on it, although it does seem like a hassle. With uhpc you'd also need this filler.

With grout you could get away with filler material but you gotta pay attention to the manufacturer specs, grouts have a maximum layer thickness, from what I've seen and I assume once you cross it stuff like performance issues/ trapped air start to happen. I'm not sure about this myself.

What say you Pete? Maximum layer thickness in grouts?

[Texasbowhunter]

A question just came into my thoughts...
I'm in the process of learning the ins and outs of working with silicone molds and from what I have learned is silicone will bond to itself after it has cured without any hesitance whatso ever...So with that said...
Will epoxy adhere to itself after it has cured without any hesitance. I ask this since working with Acrylic after I pour its suggested/recommended to apply a source of heat to the fresh poured to remove the air from the application and allow the air to rise and pop...
To ensure as much of the air is removed while its being poured or packed will the need for heat be applied so help with the air pockets trapped?
Paul

[ardenum2]

You can fill the mold with small batches of EG one by one. You have to pay attention to epoxy cure times, eg. if it cures in 9h you have 9h to fill the mold. You don't wait for one batch to cure before you add another one. It all has to be done within that 9h. Though you generally want to do it faster because as time passes it becomes thicker and harder to work with and harder for the bubbles to escape.

[Texasbowhunter]

Thank You Ardenum

[Texasbowhunter]

Peter
Just got off the phone with Sika inquiring about the Ultra and the 3350 here in the states...According to him You guys are the only ones to have this product available to the consumer...He did not mention he could get it nor did he volunteer to try and obtain the product...What he did recommend was Sika 42 Grout pack which sounds like his description was an epoxy base and has a 100 mpa...
Me Not knowing what the 100 mpa was referencing, he had a comparison to what he could find being as close to what the Ultra or 3350 was advertised as that is what he offered and suggested he in the states...
This is what he suggested: https://usa.sika.com/en/construction...grout-pak.html
Paul

[peteeng]

Hi Paul and Ard - The engineering grouts are self de-airing and should not be vibrated as this can segregate the mix. I expect this to be the same with thin casting epoxy mixes as well. I use Sika products as examples as they are internationally available. Local suppliers will have similar products.

Take Sikagrout ultra - it specs 10-100mm pour. The max pour thickness; same as epoxy is driven by the exotherm in the curing process. At some thickness say 100mm the internal heat can't get out fast enough and the exotherm can overheat and damage the cast or the expansion will cause cracking or mould distortion. Adding aggregate to epoxy and grout gives the heat somewhere to go internally then get out slower over time. Grouts and epoxies come in low exotherm versions if thick casts are required. Your local supplier can guide you on this design point. As usual follow the suppliers instructions.

Ard I have HE80 casts here now over 2 years old and using a 10x magnifying lens I can't see any surface changes or crazing. The grout people have to guarantee their products performance for 40-50 years sometimes 100years so I'm happy with the grout that if poured and cured correctly it won't microcrack. For me I get a specified product out of a bag. The self made EG path does not have the certainty of the modulus unless tested. I am having some epoxy/ALOX tested soon. I have been using epoxy for over 40 years and so I'm confident about it as well.

Regarding air pockets - 1) mixing grout or epoxy will introduce air 2) both will de air the major volumes of this introduced air. Some people do try to degas the epoxy mix under vacuum if they have the equipment. But using thin river casting resins means it will de-air to a satisfactory level by itself. And yes epoxy will stick to cured epoxy. If it can be roughened up to take the shine off it all the better. Yes grout or epoxy can be coloured with pigments but I think post painting is the go if you want colour. Especially polka dots or stripes. This will seal the surface better in both cases.

The biggest issue I see with EG is in drying the aggregate. I do vacuum casting and I see huge differences in test results of various laminates if the fibres are dry vs done in ambient conditions. I expect that the aggregate will have huge amounts of water in it and unless its heated or vacuum dried how do you get the moisture out? or take it to the desert and normalise it then cast there. Thats why there are many aerospace and composite manufacturers are out in the desert (other than the cheaper land) as the composite environment is then dry. So if you live somewhere with low RH all good. But I suggest if it's raining it's not a good day to do the pour. Epoxy and water do not like each other.

Paul it is possible with thin resins to dry pack a mould then pour in epoxy until the mould is full. This is how I would do it. To figure out how much epoxy you need you can dry pack a known volume container with the proposed media, weighing everything, then fill the pack with water until neat. Then weigh again. Now you know the epoxy weight to add to your mix for 100% pack. The epoxy I use has a density of 1050kg/m3 so is same density as water.

Do not apply local heat unless its cold in your area and then follow the suppliers curing info. Surface bubbles are cosmetic.. if you create a hot spot in epoxy it will become a heat cell and cause local expansion. This then over heats and warms the area beside it this overheats etc. This creates an uneven cure in the part... if a hot spot gets really hot it will burn. Slow initial cure is best then a post cure if possible. With epoxy, post cures improve strength and modulus usually drops a little. In machine parts strength is not an issue as its been designed for stiffness. The low viscosity casting epoxies are designed to cure slow and release heat slow.

If you like the epoxy look then laminate some chopped strand matt or aggregate to the mould first as a "tie layer" then fill with grout once cured. Like an epoxy aggregate footpath. They have similiar mechanical properties so thats an OK workflow.

Keep researching. I've been researching, building & testing now for over 2 years and have come to some conclusions for machine parts... Peter

[peteeng]

Hi Paul - The sikadur is epoxy and it has a modulus of 9GPa. Aluminium is 70Gpa and steel is 200GPa. The grouts run around 40GPa and the 3350 is 58Gpa. I have found a grout in the US for someone that's good. Its in my Milli thread I'll do a search. Peter

Heres the NA grout I found - https://www.ctscement.com/product/ul...?c=GROUTING&t= they also have other non shrink grouts that maybe suitbale.

[ardenum2]

Thanks for the moisture pointer, I live in a relatively dry area, in human terms at least, to the point I have to humidify my bedroom so I don't wake up in the middle of the night with my mouth glued shut. The air does lack the temperature of a desert though. Have to think this one through.

[peteeng]

Hi Paul - The ultraflow is a 100% CSA grout with sand added. So its a very different animal to portland concrete... Peter so if your in a dry area you need to cover the cast and keep it wet for some days to stop the surface drying out ...

[peteeng]

Hi ard - epoxy and grout dont need a lot of heat to cure the suppliers datasheet usually has the min temp for curing in it

Hi Paal - What are your reasons for cold casting a machine? Interested in your thoughts. Peter

[ardenum2]

Well whatever I search locally for in terms of high performance grout I end up at dyckerhof and uhpc, oh well...