Triangular Gantry Beam

[RaderSidetrack]

Edit: Unfortunately CNCZone heavily compressing the images

I can't do anything about that, but one work-around would be to post images to a 3rd party photohost, then post a link in your thread. There are lots of hosts out there -- but the one I use and recommend is https://postimages.org

You can upload images for free, and are not required to even have an account. A free account is not required, but does make managing images easier.

[ardenum2]

Then use Tetrium-steel or carbon for the rail mounts etc. This will allow me to route the surfaces on my router. Got to make a bigger machine so I can make small machines...

is tetrium/carbon gonna be just a precision surface with threaded inserts underneath or you'll make it thick enough to hold the bolts?

I did a simulation a while back for a copy of dmg's uPrecision and it is virtually identical to having a long monoblock metal piece. So segmenting it is viable too, at least in virtual.

Another idea I had was using a long thin 5mm metal strip with bolt holes glued on top of the cast and having threaded inserts with anchors underneath the holes, threaded inserts not connected to the strip.

[peteeng]

Hi Rader - The issue with using 3rd party is that the links time out at some point.
Ard - The bond strength between steel and grout is about 1MPa (I have to check how this is tested but I expect its a shear strength). If I epoxy the inserts in the bond strength is 30Mpa in shear. So I shall be Moulding in holes and use drop in anchors such as these but epoxy them in.

RS PRO Steel Drop In Anchor M12 x 50mm, 15mm fixing hole | RS (rs-online.com)

The Tetrium takes threads and bolts quite well so I could just make slabs of that, drill and thread, bond to grout where required and post machine. Early days. have to decide whether its a mill or router and how big. But lets get Pippin thru this gantry. Peter

[peteeng]

Hi Pippin - Here's what I did.
steel sections 200x200x10 all gantries 1000mm long corner outside rad 12mm
grout 200x200 with dia100mm hole grout is 40GPa modulus

pads on bottom are 50mm wide to create restraint face - 12mm thick but see step and fusion file attached
The z plate is rigid and the load is 300mm below the gantry. I placed 1000N on bottom edge of plate

Results
steel SHS worse at 90um
grout at 20um
filled at 9um

and -all connections are bonded. Even the fill to the steel. In reality this will be mechanically locked vs bonded.

So I also ran the grout to steel as a friction connection with a co-efficient of 0.3 and the deflection was 13um vs 9um so there is a slight slip there... Peter

just looked up steel to grout cast friction and its 0.6 so rerunning solution now... Answer is....... 13um so still a little slip at 0.6

grout gantry is 64kg and steel is 60kg so close. The 100mm hole could be made bigger to lighten the casting...

[ardenum2]

Ard - The bond strength between steel and grout is about 1MPa (I have to check how this is tested but I expect its a shear strength). If I epoxy the inserts in the bond strength is 30Mpa in shear.

If you only use inserts with a thread but no anchor you need to watch out for torsion forces not just pullout. here's a spec table for granite inserts commonly used and optimized for torsion too. (see table)

I'll be using these https://www.aliexpress.com/item/1005004529261503.html, screwing a hexbolt 5mm in and then welding it together.(hexbolt to counteract torsional forces and pullout forces at the same time).

[peteeng]

Hi Ard - Thats why I prefer to epoxy them in vs just cast in. Casting in needs a good resistant shape. Plus the insert has to be knurled or roughened. There used to be some hexagonal inserts around. Or you turn up something yourself... Peter

The Pippin Study cast gantry is 65kg the steel one is 60kg so close. To cast 65kg in Lanko is $96AUD to cast in Sika 3350 is $208AUD. If the HE80AG is similar to the bunnings Lanko price then I'd use the HE80AG (E40GPa density=2200kg/m3 )

[peteeng]

Hi All - Sika have come back with some MOE.

sikagrout 212HP 25GPa
Sikagrout Ultra 33GPa

Peter

[peteeng]

Hi All - Once you decide to go grout then you have to decide how to attach lands for the rails and bits. Grout (concrete) has a COE of 10-13x10e-6deg C. G10 fibreglass has the same range and 10-12 is the usually accepted number for G10. So its a great match to the grout. I make a similar product to the G10 and many years ago had to have it tested for strength, electrical and fire properties so know its data well. But you can buy Garalite or G10 at many plastics places from very thin to thick. Being epoxy its easy to rough it up and bond it onto the casting. Now steel has a similar COE to concrete as well. Its accepted as10-12x10e-6 etc as well. So is an excellent match. Aluminium is 23x10e-6 so expands twice as much per degree. If a tough bond is used I expect that it will be OK. I have been bonding aluminium bits to fibreglass for 20 years plus and I don't know of anything that has fallen off (yet). So I'm going to use G10 for all the hard points, they will be threaded and drilled same as metal. In this way I can machine the parts on my router.

Next thing to discuss is the mould - use white melamine for flat stuff, machined mould for curvy stuff or hand shaped MDF or ply. Look up standard mould making practice for fibreglass and follow its lead. Use paste mould wax on the melamine say 5-6 coats. Most of the poor surfaces I see and the mould stick ups are due to not having waxed the mould properly. There are oils and greases used in the concreting industry but we would aim at a top quality finish. Good Luck Peter



don't use PVA as it is water soluble! This is not PVA glue (poly vinyl acetate, correct abbreviation PVAc ) PVA release is poly vinyl alcohol a different compound. Its main claim is that it is impermeable to gases which means the styrene in resins can't attack the mould. Look up casting concrete table tops there's lots of info on that.

[pippin88]

I have been hard at work building a fleet of flying gantries (see attached image).

I've looked at triangular(ish), rectangular, semi-circular. Triangular appears to win.

I've been looking at adding expanded polystyrene (EPS) to concrete / grout - attempt to make it lighter and use bigger sections. But there are issues - mixing (EPS tends to separate) and most research shows a greater reduction in modulus and strength than the reduction in density. If you halve the density, you lose more than half the modulus. Adding carbon fibre strands appears to help a lot.
But then we are in the area of experimental mixes / poor reliability - see the issues with epoxy granite.

At present a triangular(ish) gantry made of full strength grout with large voids in the middle (with ribs in between) appears the best solution.

[peteeng]

Hi Pippin - They seem to be in a rush to get somewhere, maybe to join with your ducks on the wall? I too have been busy with the FE and have decided that 150x10mm aluminium flat bar with plywood centre is the go for the next gantry. [150x10 al / 150x(5x16mm) plywood / 150x10mm al] to make it 150mm wide and 100mm deep. I am going to cast a self levelling table from grout to make a level bench for laminating gantries and plates on... Peter

[ardenum2]

Is this just an experiment or are you building something? how big is it going to be? I've done a lot of research using cores and they become more viable the larger your parts are. For all my parts, which are relatively small, the weight reduction was marginal, but the work associated with using the cores was not.

Remember to use minimum wall thickness of 50mm for all structural walls. This number comes from years of uhpc research at durcrete/dyckerhoff. I take that at face value.

[pippin88]

Ard, both - thread partly started to explore the theory, but mainly to build a mill.

I picked up a great base cheap. Old optical inspection machine base. Granite in a U shape i.e. high wall design. 600kg of precision granite that I picked for free (well had to hire a ute to get it home).

Walls are 300mm high (front top surface of the base plate). 670mm between the walls.

So I have need of a gantry about 1m long.

I am wanting to do this right (have already built a few routers, each better than the one before but still not satisfied). Servos. Ground ball screws. Etc

[ardenum2]

wow, that is a magnificent piece! definitely worth putting time into. 2 rails per side, this is gonna be epic.

Let's do this!

Start with the spindle first. What do you think about this one:

https://www.ctbservo.com/product/mot...11-2500-12000/

it was $4500 in 2020.

upload the cad of this bed once you put it in. Let's all have some fun with this one.

sad for those stators though...

[pippin88]

Unfortunately I missed about by an hour or so on the rest of the machine. It had linear motors and dual lear rails on each side. Someone got in first and stripped the linear components.

I have 45mm Bosch Rexeoth roller rails. Way overkill / sized, but I got them cheap new old stock. They are too big for the Z, so I'm looking at 25mm roller rails for Z.

I had been planning a lifting gantry design, but when this popped up it was too good to let pass and so the design must change.

[ardenum2]

These linear motors would not work for you, the stators are too narrow so the forcers would not have had enough force to move a milling gantry or they'd be too long, you'd loose on the travel. Send a text to that guy and tell him you want ransom for the stators...

check these guys: https://tnkbearing.en.alibaba.com/ they got hiwins and rexroths too, ina as well unless you wanna snipe them on ebay. Size 45 that's big, but should work well for a heavy Z axis if you mount them on the gantry.

[pippin88]

I figured the linear motors would not be useful, but the rails and the gantry would have been nice. Still a very good score to get this base. Saved me a lot of time and $.

[pippin88]

At present I am trying to work out how to do the gantry connection to the Y axis linear bearings.

The problem is that I can't cast a gantry with enough precision for length

The ideal connection (in theory) I think is to have the gantry resting on top of the linear bearing and bolted down, and to have a side plate as well.

I need the gantry offset backwards partially on the linear bearings or I lose too much travel.

[ardenum2]

upload the bed in step. I have a few ideas. those 45's are huge indeed. Will you be mounting the gantry rails directly onto the cast or will you cast in metal blocks and have them machined? That gantry is very healthy by the way

[peteeng]

Hi Pippin - Your in a tough place; small machine, big bearings, very tight space for the saddle. My current solution for this is to place the gantry rails on top and make a 90deg angle saddle. This separates the bearings for easy bolt access and moves the Z axis back as close as possible to the gantry... Make the gantry as shallow as possible to minimise leverage and I find it works very well. Peter

[pippin88]

Peter, I think that would be worse for my machine. I need to offset the gantry backward partially off the bearings.

ardenum2 - undecided re mounting as yet. Practicality of rail bed preparation / machining is going to impact on the design as well...
Attached the base (and Bosch 45mm rail models) in a zip