[peteeng]

Morning All - Now I'm up to Milli-Block-14. I have run it in several "rigid" forms deciding what the weak link is. By running it with a rigid gantry for instance, tells me if that makes a big difference or not. The biggest jump is making the columns rigid. From 24 to 44N/um. The next is the base from 24 to 37.7. So it seems this area could be better. So I'll make the columns wider and narrower (to keep to 50kg) and see what happens. I had an idea that a steel frame would provide a sort of two class machine. Without the frame 20ish and with the frame 30ish stiffness. But it contributed very little. Even solid Aluminium didn't do much... You would think that a 170mm solid square piece of aluminium would be a great column!! Seems a good thing to up size the rails as well... and I've made the bridge bigger...Peter

[ardenum]

How about reinforcing with threaded rod and maybe even slightly pre- or post-stressing it? You'd need to make the bridge in the middle for that though, unless you'd only do it with the column-bed joint.

[peteeng]

Hi Ard - The outside skin of a part is doing most of the work. Look up second moment of area

Pre or post tensioning does not improve the structural rigidity. Tensioning is used to 1) remove slack in a system eg a spoked bike wheel 2) to move the average stresses to the compression side so when in tension it does not fail eg a concrete beam, concrete is poor in tension so move the internal stress to the compression side so when it gets loaded its always in compression. These strategies do not improve rigidity only stability or strength... Thanks for thinking about the issue, all input appreciated. Peter

[peteeng]

Evening All - Just been back from town getting a Pizza. I have started a new Milli series called Milli-Gram. I widened the columns as this seemed to be the thing to do. It only raised the X stiffness from 24 to 27 N/um. So I think were clearly in the area of least returns. To understand this a little better I drew up the load moments of Milli. Turns out the X axis moment is twice the other moments so Milli has to be twice as stiff in the X direction as the Y and Z dirn. This is the same as other mills probably worse due to the extra stackout that would be in the Z axis...

As you increase the X axis stiffness the other axes naturally get stiffer. As I'm trying to minimise weight I hit limits as I can't just make everything bigger and bigger. So I calculated the equivalent stiffness column in steel and that turns out to be half the weight of the CSA one. So a quandary is upon me , do I go steel for the columns???

I'll have a little think.... Peter just noticed my calc is wrong I'll fix it in the morning...

[Zorbit]

Hi Peter,

Does your design lend itself to casting the locating surfaces for the X and Z axes together ? It's crucial that they are correct to each other, perhaps one precision made jig could cast the reference surfaces for these two, and once those rails are fitted a second jig could cast the surfaces for the gantry rails, referenced from the previously installed rails.

It would mean a second and third casting process, perhaps with something like a slideway repair compound that can be injected.

[peteeng]

Hi Zorbit - The current plan is to have aluminium inserts in the cast CSA parts that are post machined so all part fits are precise. This includes the rail surfaces and car mount surfaces. Haven't got to that yet!! The inserts will either be imbedded in the cast or post epoxied on, in suitable pockets. Once I get familiar with casting the CSA I may find that its surface is good enough to cast precise registrations on directly. That's down the track a bit.... Peter

and here's the image to correct the typo in the prior post. Doesn't change the outcome...

[pippin88]

Peter, seems no surprise

A stiffer material in larger hollow section size will always have the advantage....

A solid section of aluminium is not optimal use of the material

5mm thick steel tube will ring like buggery and dynamic performance may be quite bad despite reasonable static stiffness.

I have a 200x200x6mm steel tube for my gantry on my router. It screams. Don't think it would be good for a mill.

[ardenum]

Peter, seems no surprise

A stiffer material in larger hollow section size will always have the advantage....

A solid section of aluminium is not optimal use of the material

5mm thick steel tube will ring like buggery and dynamic performance may be quite bad despite reasonable static stiffness.

I have a 200x200x6mm steel tube for my gantry on my router. It screams. Don't think it would be good for a mill.

Is your tube filled with anything? I was planning to use 250x250x16 tubes filled with EG for my vertical columns.

[peteeng]

Hi Pippin - Yes no surprise, just trying to get an insight into a answer. The material is CSA although I use aluminium in the FE model as its the target stiffness and density. I definitely would not use 5-6mm material as it would ring as you have found... Steel would have to be 12 or 16mm thick then I'd fill with 400kg foam. I wouldn't use a SHS or RHS std tube, I'd have laser cut bits made then braze together. So I can control dimension and thicknesses better. Speaking of brazing here's a nice video on it. Only found out through this video last week that you can get in-line fluxing, brilliant!! Been brazing for 30 years and never heard of it...

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

https://www.youtube.com/watch?v=fWF0...ature=youtu.be

Peter

[catahoula]

Pete, have you tried brazing any heavy wall steel? I haven't, but having just spent a significant amount of time just preheating 12mm steel to ~200 C with a medium size cutting torch on full blast (went through 3 tank refills to build my frame), I can't even fathom how much gas or how much time it would take to get that stuff up to brazing temp with a brazing tip, let alone the whole structure evenly so it's not warping. Also, the residual stress from the original rolling of the steel would still be there in most parts of the frame, so distortion while machining is still going to be an issue. For heavy steel I don't see how it's a viable option.

[peteeng]

Hi Cat - 6mm is about the thickest I've done. Its like torch welding you use a hard cone and heat up a small area say 30mm diameter then wet that then move along. You tack it out like normal welding then braze across the tacks.... Done a lot of aluminium brazing up to 10mm and that needs a lot of heat... You can use high temp solder it melts at 200degC Its strong enough for benchs and machine parts but it will melt if stress relieved same as brazing no good for thermal stress relief... Peter

Theres flow brazing and weld brazing
flow is heating up a large area and letting the rod flow through the joint; welding is like TIG small heated area dob of braze move to next dob...

[pippin88]

Is your tube filled with anything? I was planning to use 250x250x16 tubes filled with EG for my vertical columns.

Not filled

Wish I had or had designed it so I could easily fill.


16mm thick is a different kettle of fish than 6mm.

[ardenum]

Not filled

Wish I had or had designed it so I could easily fill.


16mm thick is a different kettle of fish than 6mm.

you could drill holes from the outside, put bolts in also from the outside so they stick considerably into the inside and then fill it with EG.

[spumco]

Pete, have you tried brazing any heavy wall steel? I haven't, but having just spent a significant amount of time just preheating 12mm steel to ~200 C with a medium size cutting torch on full blast (went through 3 tank refills to build my frame), I can't even fathom how much gas or how much time it would take to get that stuff up to brazing temp with a brazing tip, let alone the whole structure evenly so it's not warping. Also, the residual stress from the original rolling of the steel would still be there in most parts of the frame, so distortion while machining is still going to be an issue. For heavy steel I don't see how it's a viable option.

[Forgive me if I'm telling you something you already know, but we've finally touched on a subject where I have some expertise]

As Peter mentioned, there is brazing, and braze welding. Two different processes, though they are referred to in somewhat interchangable terms. I believe Peter was suggesting that a large steel structure could be assembled using braze welding, not brazing.

Both processes use flux, and both are done at temperatures below the melting temperature of the parent/base material.

Brazing uses a silver-based alloy which is drawn in to a tight-fitting joint by capillary attraction. Like soldering pipe joints, only the filler melts >800F. Heat is applied using fuel-based torches or the parts are heated in an oven (induction usually) with the filler & flux pre-placed in the joint. Induction brazing is extremely common in mass-production.

Braze welding uses a brass-based alloy. The filler is deposited on the surface of the base material or in joint with geometry much like fusion welding. There is no (or very, very little) melting of the base material. Heat can be applied using fuel torches or with a GTAW (arc) torch.

Brazing is extremely strong, but requires the entire joint to be heated to the correct temperature. I believe this is what you were referring to when questioning the viability of brazing a large structure.

Braze welding varies in strength depending on the filler material. It can be fairly soft (building up a damaged shaft with brass) or it can be extremely hard/strong (joining two components with aluminum bronze)


You are correct with your concern about thermal distortion, but that's not the most significant source of post-process warpage. The most significant source of warpage and residual stress comes from weld filler shrinkage when fusion welding is performed. This is where braze welding shines: the filler material still shrinks, but generally does not result in the same amount of residual stress (and resulting warpage) as with fusion welding because the filler is more ductile during the solidification process.

In addition, if one were to use hot-rolled steel there is much less residual stress from the original rolling/forming operations. Cold-rolled steel has a massive amount of residual stress - I'm sure you've seen a plate turn in to a bannana when you machine off one surface. Not so much with HRS.

Aside from filler expense, a braze-welded steel structure would be incredibly strong and have much less distortion than if it were fusion welded. The only caveat would be to keep the service temperature below ~450F - not a problem for a gantry.

And yes, you can braze very large items successfully - but it takes a significant amount of fuel & effort. I have personally brazed a 36" diameter cast bronze valve to 70/30 copper nickel pipe a few times. Imagine 4 guys with giant rosebuds all walking around the valve with two more guys standing by - one with about $5k of silver filler rod ready to jump in and stab the joint at the right time, and one with a big hammer to hit the valve to 'encourage' the pre-placed ring of filler to wet out.

Hot and stressfull job, that.

-R

[peteeng]

Morning All - Harris Inc has a lot of info about brazing on their site: https://www.harrisproductsgroup.com/...Documents.aspx Maybe soon we can just glue the joints together with something like toothpaste... I'm sure we can now... Here's a shot of a joint that is created using wrapped fibreglass. I did not do this but I have done similar work on skiff tramps. There's lots of ways to do these connections.... Peter

[catahoula]

Pete & Spumco- Thanks for the clarification between brazing and braze welding. I have actually done a bit of both styles, but was kind of lumping them together in my mind. Clearly with a sharp and hot enough torch and localized heating it would be much more manageable. My experience with braze welding also included preheating the entire joint to some degree to make the fillet-building go quicker, but that may have been less necessary.

So you would be confident machining a braze welded hot rolled steel structure and expecting it to be stable?

Helping the cause in the context of machine tools is the fact that the structures should be of a fairly stable shape already, if they're going to be rigid in their final application in the machine. Yet another reason to go larger in form.

5k of silver solder in one shot...wow!

Edit to add: you mentioned GTAW torches (TIG brazing)- I'd forgotten about that; might be great for this application. I have seen some significant spring action when cutting up HRS tubes, but maybe as long as the machining is limited to just some light surfacing on a thick structure, it would work out fine.

[peteeng]

Hi Cat - In terms of stability yes, but it can only be as stable as the materials its made from. But I think for large structures UHPC or mineral casting is the future (and for many commercial ops its already the case) for DIY and professional machine parts. Peter

[peteeng]

Hi Cat and others - re fillet building - To create equal "strength" joints to welding, a large fillet is required. But for a machine part the strength is not the issue its stiffness which can be achieved using a small fillet or even if there is no fillet (soft solder - silver solder) but the gap is filled with braze then you have nearly 100% local stiffness equality which is what you want. Bronzes have similiar stiffness to cast iron 100-120GPa so they have excellent strain transfer in a joint. cheers Peter

There's also MIG brazing which maybe easier for many people--- would help those that can't get to thermal stress relief...As this process does not melt the parent the resulting internal strain is considerably less then welding... and TIG brazing as well...

https://www.millerwelds.com/resource...brazing-basics

The large fillet is required as the yield strength is about 1/2 the strength of the parent but as stated this is not a strength application. In Milli the max part stress is about 2MPa (290psi) so its tiny tiny compared to the yield of the materials used that's why CSA concrete can be used as its tensile strength is about 10Mpa...

[peteeng]

Afternoon All & others - Setting up a mould to cast my first CSA mix. Will be a 3.5kg block. Sorting some plastic sheeting and mixing stuff... I'm also trying to repair the thermostat from the Argus laser. My guess is that it has a rust pin hole in the bottom and "died" when the water got in. Hopefully by drying it out under vacuum it will "work" again. Its been in vacuum under 500Pa for 3 hours so that should dry it quite well.... lasers been in storage for 2 years so things have dusted up or rusted up. Slowly working through its issues....Peter

[peteeng]

Evening all and sundry - I've done a small optimisation on E70 material vs steel. A solid column like I have now weighs about 78kg/m whereas a 12mm thick steel one would be 54kg/m.. I made a model and compared the two and they are about right. Both are over the 20N/um target so thats OK. The steel by the time I add a thick flange comes in at 52kg and the E70 material is 56kg.. So again I think I'm at a balanced point. If steel I'd prefer the front face to be 16mm to allow for machining.... decide after some casting tests to see where I get to with that. If I get to E80 then its a no brainer for the concrete casting.... still looking for a tabular alumina supplier... Peter