[peteeng]

Hi - If you look up Ramset or similar companies that sell mechanical anchors for concrete they have several threaded inserts with rough externals for this application. They also have design guides for strength and position. There are also epoxy set types and they have good design info for. Loxin is another.

https://ramset.com.au/Product/Detail/29/Threaded-Insert

The bond strength of grout to steel is very low so its important to use an odd shaped anchor to mechanically lock the anchor in place, its not really "bonded" in. I prefer to drill and epoxy the anchor in then its properly bonded in with little internal stress. Cast in anchors can eventually get cracks propagating from them due to the stress created in the plastic shrinkage stage of the cast especially if too close to a surface. I have been involved with lifting and holding down metal forming equipment in concrete floors and the epoxy type does better then the mechanical type. Its quite a fright when large chunks of concrete come out of a floor for instance when a few tonnes of machine comes loose and starts dancing.... Peter

[NordicCnc]

For anchors I'm doing these https://www.aliexpress.com/item/1005004529261503.html screwing in a hexagon bolt 5mm in and then welding it around the thread. Threaded inserts m4 20mm, m6 25mm, m8 30mm, m10 35mm. Total length with a hexbolt 50mm for m4, 60mm for m6, 70mm for m8 and 80mm for m10. Cost is about $400 for a total of 300 inserts. Unless you can make them yourself, like in the picture attached.

The slides have english in them, but not everything is translated. What kind of a machine are you building?

I have the possibility to make them myaelf since I have a lathe.

I am not building yet, I am still designing it. In general it will be a 6040 fixed gantry moving table machine.

Skickat från min SM-A528B via Tapatalk

[suspension]

I have seen many DIY machines have a steel or aluminium plate between the T-table and X axis rail blocks. Is this really necessary?

Also, can we use epoxy that are advertised for jewelery, lamination and flooring for this purpose?

[NordicCnc]

I have seen many DIY machines have a steel or aluminium plate between the T-table and X axis rail blocks. Is this really necessary?

It makes aligment easier when you can first align the rails. Then you can bolt your table onto the intermediate plate and shim that / surface it so that it is aligned with the rest of the machine. You need to design a machine that is adjustable so that you can fine tune the alignment. Once one parts is aligned, you are able to "lock" that part and move on to the next, without affecting already done alignment. One example is to use an intermediate plate between the gantry axis and the spindle axis. That allows easier tramming.

Skickat från min SM-A528B via Tapatalk

[ardenum2]

Also, can we use epoxy that are advertised for jewelery, lamination and flooring for this purpose?

any low viscosity epoxy should work, anything <500

[peteeng]

Hi Sus - I think what you are seeing is a spacer. If you use a flat plate for the table you will need one. This is the same for other drives as well.... There is a difference in stack height between drive nuts and bearing cars. So spacers or a machined gutter in the table is required. In regard to epoxies use 100% solids resin. Flooring epoxy for instance will have a solvent in it that will evaporate and allow the resin mass to shrink. Laminating epoxies are good but have thixotropes in them to prevent sagging. If you are talking about a resin suitable for making mineral cast parts then use an epoxy suitable for river table casting or infusion. These are thin, clear and have no thixotropes. Peter

[ardenum2]

There is a difference in stack height between drive nuts and bearing cars

and a difference between all of that and bearing support on both sides, which is why FK FF supports are much better, you don't need to use shims or epoxy, of course the mounting surfaces still need to be parallel to each other, since all mounting surfaces are vertical and parallel to each other you have a few mm to properly adjust everything in the vertical plane. All the components are then flange mounted like the nut. A ZKLF bearing would be even better than FK/FF's.

[peteeng]

Hi Sus - If you are casting parts then you have the opportunity to cast in flanges or lands to have correct mounting hts for bearings. Time to get busy in CAD. Peter

[suspension]

It makes aligment easier when you can first align the rails. Then you can bolt your table onto the intermediate plate and shim that / surface it so that it is aligned with the rest of the machine. You need to design a machine that is adjustable so that you can fine tune the alignment. Once one parts is aligned, you are able to "lock" that part and move on to the next, without affecting already done alignment. One example is to use an intermediate plate between the gantry axis and the spindle axis. That allows easier tramming.

Skickat från min SM-A528B via Tapatalk

Sorry I dont understand.
Once I get two linear rails parallel to each other, their bearing blocks will also be parallel and their top surface will be at the exact same height. If I fix a T-table which has precision top and bottom surfaces on rails blocks, I will get a top surface which is parallel to rails. Not sure why I need a plate in between these two in this case.

[NordicCnc]

Sorry I dont understand.
Once I get two linear rails parallel to each other, their bearing blocks will also be parallel and their top surface will be at the exact same height. If I fix a T-table which has precision top and bottom surfaces on rails blocks, I will get a top surface which is parallel to rails. Not sure why I need a plate in between these two in this case.

Then you can skip the intermediate plate.

Skickat från min SM-A528B via Tapatalk

[suspension]

. In regard to epoxies use 100% solids resin. Flooring epoxy for instance will have a solvent in it that will evaporate and allow the resin mass to shrink. Laminating epoxies are good but have thixotropes in them to prevent sagging. If you are talking about a resin suitable for making mineral cast parts then use an epoxy suitable for river table casting or infusion. These are thin, clear and have no thixotropes. Peter

Hi Pete,

I found this in UK and its data sheet is here. Would you be able to have a look and let me know whet you think?

Thanks
Sus

[NordicCnc]

Hi Pete,

I found this in UK and its data sheet is here. Would you be able to have a look and let me know whet you think?

Thanks
Sus

45min pot life is too short, you will kot be able to mix it properly and then cast. Aim for at least 3h pot life.

Skickat från min SM-A528B via Tapatalk

[suspension]

Hi all,

CNC machining does not seem like a viable option. I tried to get quotes from few workshops, but many refused, some said their machines are not big enough, one quoted and said they can't maintain tolerances of .01 for such big work pieces.
I think I have now two options, one is to get it milled anyway but test for accuracy and if that's not enough (probable) do the scraping as this guy. This would require lots of patience and expensive measuring tools which I might not need.
Second option is to use epoxy for leveling surfaces. For example DWH 310FL. That was the approach used by Stef's build which I am basing my machine on. According to him it worked, but not sure how things are after 1000 hours of use.

My current preference is to use leveling epoxy. I have two things to get clarified.

1. If I am not putting this to a CNC mill, I cannot mill holes and threads required for linear rails, motor brackets and ball screw bearings. I am planning to use a simple manual CNC milling machine for drills and tapping. This will be done individually for each steel plate. Also, I can use the a manual mill to create a register which can be used to align the master rail. I hope this will work and the accuracy I get is good enough?

2. To be able to get parallel surfaces on top of all steel plates on a given granite/concrete base, I will have to make sure to connect all areas that will be filled with epoxy so epoxy will have a free flow between different steel plate areas. To keep this simple, I may have to change the design so all plates are at the same height. In addition, I will have to cover all threaded holes with something (eg: Silica glue or kids clay?) so epoxy wont go inside.

Also, to reduce the thickness of epoxy layer required to get same level, I can surface mill (or grind) each steel plate individually. During mold making, these will rest on MDF (25mm thick or 32mm thick) so they will approximately be aligned.

What are your thoughts about this process?

Thanks
Sus

[joeavaerage]

Hi,

My current preference is to use leveling epoxy. I have two things to get clarified.

Cured epoxy has a Youngs modulus of somewhere of 3Gpa to about 6Gpa, in short is 'f****king squishy' and when the load comes on it will deform.

CNC machining does not seem like a viable option. I tried to get quotes from few workshops, but many refused, some said their machines are not big enough, one quoted and said they can't maintain tolerances of .01 for such big work pieces.

Tough s****t!! You and every other hobbyist has faced exactly the same issue. You just have to find a company whom has big enough equipment of the quality to meet the requirements.
It will COST UNTOLD!!!!!

The company that did my machine (cast iron axis beds) had one near new (3 years old) Okuma double column mill that you could park a car in, with a 75hp spindle, three million dollars worth.
The machine that they used to do my axis beds was much smaller, another Okuma about five years old and only $800,000!!!

To say they charged me plenty is an understatement, it cost untold. But the job is perfect, I got to meet the machine operator and he did his best for me. The foreman whom did the toolpath
programming is very nice, it was he that showed me all around the factory, and the boss, well he just took my money, efficient as you like. LOL.

Finding a machining company with a capability and interest to do this sort of demanding work for a hobbyist is a challenge in itself. Keep looking.

Craig

[peteeng]

Hi Sus - You are getting to the pointy end of whats called the "fuzzy front end" of a project. If you can't have a machine shop do the work then its highly unlikely that your machine will achieve the accuracy that you want. Levelling with epoxy is not recommended. Sure you get a level surface but epoxies modulus is like rubber and you need 6mm of it to self level. less than that and it will crawl and not be self level. If the resin is really slow curing and thin you maybe able to do 2-3mm but do you want 3mm of rubber under your rails? I think you need to keep looking and find a suitable machinist.

Now it is possible to hand make the surfaces via lapping and scrapping and thats how the machine makers 200 years ago did it. But it takes a lot of time and patience to achieve a good result.

Look up lapping using the 3 plate method.

Re - epoxy you linked. It has a viscosity of 400cps which is quite thick. It will pour its a laminating epoxy so is thixed. Which means if you agitate it it will thicken. Look up Non-Newtonian liquid. This prevents the resin running down hill when laminating.... which also means it will not fill in your mixture really well. Also note its modulus is 2800MPa which is very rubbery. ie if you used this for levelling your mounting on rubber.

https://www.youtube.com/watch?v=kCwNN1DLePs

https://ericweinhoffer.com/blog/2017...-plates-method

You start by making (buying) a suitable flat plate from granite or stabilised grout. Plus make suitable laps for each land on each main part. Then you lap each land in so it is square and parallel and level. You will also need an engineers level and lots of engineers blue. A slow but doable process as our forbearer engineers did.

The other approach is to make a mill as best you can with what you have and can get. Then use that machine to build a better machine, then the next machine to build a very good machine. There are examples on the forum of people doing this. They build a plywood machine that builds an aluminium machine that builds a good mill. Again that's how our forebearers did it and we still do this today. We use toolmaker grade machines to make industrial machines and master builders make better and better machines as time goes on. Even multi million $$$ machines are still scraped into spec. Keep at it, there are no short cuts. Peter

https://www.britannica.com/biography...tworth-Baronet

[ardenum2]

badhabits Epoxy Resin L is <800cP(mPa-s), mixed with hardener GL 2 ends up at 250cP(mPa-s), 210min pot life

suss linked resin is 2000cP(mPa-s), mixed with the hardener(60 cP) ends up at 400 cP(mPa-s), 45min pot life.

sus wants to use DWH 310 under the rails not pure epoxy. also that scraping alex did is not correct, you need to do short strokes.

[suspension]

Thanks Craig, Pete and Ard for the encouragement and valuable information. I will be talking to more machine shops and see how it will go. Also will go through the video to understand better how to do this by my self if CNC milling path becomes impossible.

Also what are your thoughts on the T-slot table from this chinese vendor?

https://hwjxzz.en.alibaba.com/minisi...=1600385796218

They quoted 550 for cast iron (800 X 400 X 50 mm) and 600 for steel (EXW price). When I asked about the precision, they said its .02 for 2000mm and .03 for 3000mm! I doubt these values are correct. But in the product listening here they mention about various precision classes and grounding finishes.

Which type of T-table suits my needs better? Cast iron or Steel?

[peteeng]

Hi Sus - At 165euro per kg DWH 310 is very expensive. I pay $20AUD for a very good epoxy and add steel fibres myself. I doubt DWH310 is self levelling due to the fiber addition. That sort of product is great for its application. I have used all sorts of metal and lubricant additives in epoxy for all sorts of reasons. Best plan is machining though for a mill build. It's still only E=8700MPa so its a stiff filler. Good plywood is stiffer MDF is 4500MPa.... Peter

[peteeng]

Hi all - I just looked up the age of granites. Even a young piece is 500 million years old. We are such a pimple on a pimple.... So its had a good time to become stable. Peter

[joeavaerage]

Hi Sus,
you have run into exactly the same brick wall that many others have faced. You have decided you want 600mm of X travel requiring either an axis bed of say 1100mm long or
alternately a table of 1100mm long. Such a table or bed will require machining on a machine with say 1200mm of travel.

What's the bet within 50 miles of where you live there are 100 CNC machines with 600mm of travel but only 5 of 1200mm of travel. Instead of paying 75GBP/hr now your looking
at 200GBP/hr.

You desired machine size does not match your budget, in fact I think your $10,000 budget would be consumed three times over with a machine of the size that you want.

I faced exactly the same problem. What I did to minimise the cost while retaining the best possible size I could afford and the accuracy and rigidity required was to make three axis beds.
They are all identical, that meant that I only had to get one pattern made. It also meant that the machining tool paths were shared amongst three beds. The only difference between the X axis
and the Y and Z axes are the T slots.

At the time that I built this machine the local heat treatment facility had an oven capacity of 700mm, so my axis beds are 695mm long.

My machine then is a frame, made out of two L shaped 32mm med tensile profile cut steel sides, because I could not afford to cast it. The steel frame does not require machining as the beds are bolted by shims
at each corner to the frame. Ideally the two L shaped sides of the frame would have webs welded in between them, effectively forming a rectangular tube. I'm a pretty fair welder and
I used to work for a company that hires welding equipment. I would be using 2.4mm ESAB CO₂ shielded flux cored wire at around 25V and 350A. Such a weld has 8mm -10mm
penetration in one pass. Thus I could weld 10mm webs with complete authority. The downside is that I would have had to send the welded frame to Auckland (1500km away and on another island)
to be stress relieved at $6.00NZD/kg (2020 price, who knows how much it is now), and the frame is 210kg. At the time my CNC budget was exhausted so the best I could do was bolt the two sides
together with 20mm plates separating them. I lament the loss of rigidity, but I can always revisit this.

The issue is now that I use my machine daily to make a living so I cannot afford to take it out of service for a month while I weld the webs in place and send it to Auckland and back, and I have no more
budget now than I did then, just quietly.

The important point here is that the axis beds remain the same. I had them cast at considerable expense, and then machined at double the cost of casting......and they are good, and that was
always the intention. The frame I could alter or even replace over the years but the axis beds would be the heart of the machine and remain the same.

You may have heard the expression 'cut your jib to suit your cloth', and it applies in hobby CNC bigtime.

Craig