[keen]

Z Column stiffness is the first weakness. I measured the flex - see pics, and improved the column stiffness with a back plate over the cut out weak point.

More power without matching rigidity is likely to result in rapid cutter breakdown.

This is not the only machine weak point - it is a light mill. So don't expect massive improvements from this mod.

Keen

[johnh]

Yep, I'm in uncharted territory, maybe Tormach has already experimented with the limits of their machine frame but they haven't shared the results. In my experience it has been the motors HP that is the limiting factor not the rigidity. But you are right in your situation, everyone's machine methods and jobs are different. I mostly do aluminum.

jh

[daniellyall]

if you manly do aluminum why don't you just add a high speed spindle or a big BLDC motor

[johnh]

big BLDC motor?

[johnh]

Z Column stiffness is the first weakness. I measured the flex - see pics, and improved the column stiffness with a back plate over the cut out weak point.

More power without matching rigidity is likely to result in rapid cutter breakdown.

This is not the only machine weak point - it is a light mill. So don't expect massive improvements from this mod.

Keen

You are absolutely right. That is why I'm looking at an epoxy granite fill

What did you measure for deflection under what loading? where was it loaded?

[tmarks11]

big BLDC motor?

Brushless DC Motor.

Much better low speed torque characteristics, smoother without the vibrations you see in (cheaper) single phase induction motors.

But more expensive (in the 3-5 HP range). And requires a dedicated BLDC controller, also expensive (when talking about the 3-5 hp range).

Not as plentiful as 3 phase motors (which you can find in killer deals on eBay), and since they are not as plentiful, chances are you will end up having to make an adaptor plate as you might not find the exact motor frame you are trying to replace.

[keen]

You are absolutely right. That is why I'm looking at an epoxy granite fill

What did you measure for deflection under what loading? where was it loaded?

Hi John. You can see from the pics where the load was applied - column twisting ex the torque wrench. And see the pic where it was measured before and after. When I loosened the back plate the flex was more - but I did not do any maths on it - just the practical simple stuff.

I cant see how epoxy granite fill can do a lot to fix the main weak links - it may only improve things a few %

The base bonded to a solid base sub plate or frame with bracing from it to the column would add a lot of stiffness - but that is bulky and a lot of work and cost. And even then the XY system and head are still light - it is basically a fairly balanced light mill and improvements will be limited. Keen

[G59]

John, don't waste your time and money on epoxy granite. It has too little resistance to twist. The modulus was explained a few posts back and in a twisting situation, it is fairly weak. Good at dampening vibrations maybe, but it requires a fixed inner structure to resist bending and twisting.

[R.DesJardin]

I'm not sure if column twisting is comparable to induced cutting forces, that being said the backing plate is a decent idea just out of principle for sure. I Am following with anticipation and until you take cuts only then you will know the answer of PCNC rigidity. I love modifying stuff. So what is your final target max rpm?

[philbur]

The dynamic stiffness of the frame improves with RPM, so produce the 5 hp at higher rpms, with a corresponding reduction in cutter diameter and increase in feed rate.

Phil

[johnh]

Since I mostly do aluminum I will not be torquing the frame much with high cutting pressures and interrupted cuts as on steel with a large face mill, but with the right cutter and higher rpms steel removal will be easier.

Right now Tormach gets a multiple of 2.39 over the base motor rpm if I remember right. Using that as a rough estimate to shoot for, my motor is rated at 5hp and a normal rpm of 2875 so I think I can at least double that or 120 hz for 5750 rpm, so 2.4 x 60 = 144hz 2875 x 2.4 = 6900 rpm,

I will be running 1:1 gearing, really low rpm is only for tapping, boring, reaming with minimal power needed for those which should not be a problem to reach with the VFD so my working range I'm hoping to get 300-6500 rpm, full 5hp at 3000,so with good endmills, coolant and cutting paths steel should cut nicely.

this high rpm will really help with smaller diameter cutters, another note, for all of you that are freaking out about 5hp on a Tormach, at 120hz the HP is about half or 2.5hp, so its no wonder that a good geometry carbide 3/8 endmill with good coolant and a full width .200 deep cut at 3750 rpm is on the verge of bogging the stock Tormach motor down, that is it's limit, mine any way and that is because it's really only putting out .75 hp because of the over speeding apparently. That's what I don't think people are realizing, the machine is made for 1.5 hp but isn't being taxed to that extent, so what sounds like way overkill with 5 is really just 2.5 which equates to about 66% over the original design point.

So I'm pretty excited about the potential. I have come to my last iteration on the new drive design and I'm ready to fabricate the components that leave me some back doors and an upgrade path for my new universal spindle which is the second project. The thing that's holding me back is in the meantime its computer doesn't want to boot anymore, its an old Pentium 4 machine and the caps are leaking and bulging on the MOBO, so I have a new box coming, and that doesn't let me move the machine right now and I need to pull off the base plate so I can measure the bolt pattern. That's the last think I need for the design. Hopefully these holes are the same for the series 2 and 3, right now I have no way to verify that so right now it will only work on series 1's for sure.

this is what a series 1 looks like after 1000's of hours of sucking in the coolant mist and slinging the condensed oil around.
Attachment 276398

my new design has a separate ac fan to cool the motor in low rpm high torque mode, (it will run all the time). it will have an 8" flex HVAC duct attached to the outside fresh air, this will also keep the drive housing under positive pressure to help keep chips out.
Attachment 276400

this will be it official moniker going forward
Attachment 276402

Also the theory behind the epoxy fill is that it in itself is not that stiff, but the stabilizing of the casting shell is to let the shell work more effectually by reducing the buckling under compression. Same structural theory that lets 2lb foam core make two thin sheets of carbon way stiffer than they would be without a core in the sandwich.

I haven't loaded the model in SW finite analysis yet but the weak area looks like it would be the base right in front of the column. The easiest thing that makes the most sense is to fill the z head cavity with epoxy lead shot for mass damping right where the vibes are created. Comments?

jh

[G59]

Same structural theory that lets 2lb foam core make two thin sheets of carbon way stiffer than they would be without a core in the sandwich.

Under compression, Yes you are right.

for all of you that are freaking out about 5hp on a Tormach, at 120hz the HP is about half or 2.5hp,

Where did you come up with your horse power derated figures?

Just because the VFD is derated when operated under split phase, it doesn't mean less power to the motor when you have already compensated for it, by installing a bigger VFD.

Here's something I know for a fact, the stored kinetic energy in the 5HP motor is now double what it normally is, due to the increased operating rpm, no matter what frequency you are putting through it.

Does that mean anything to you?

Were not just a bunch of nay sayers John.

Like I said before, it's your machine. It's paid for. Nothing wrong with that. You like to tinker and more power to you.
I too want to see the end results. Don't get annoyed just because some of us are getting you to look at it from a different perspective.

Here, from Reliance Electrical.

If the motor is a 50Hz unit and you're going to be using it in 60Hz-land it will spin 20% faster.
Horsepower(hp) is proportional to Torque times RPM. Since the motor's torque is not going to change appreciably with an increase in frequency it will now provide 20% more hp. Your 8hp motor just got promoted to being a 10hp motor. Something for almost nothing!

[philbur]

Also the theory behind the epoxy fill is that it in itself is not that stiff, but the stabilizing of the casting shell is to let the shell work more effectually by reducing the buckling under compression. Same structural theory that lets 2lb foam core make two thin sheets of carbon way stiffer than they would be without a core in the sandwich.

I haven't loaded the model in SW finite analysis yet but the weak area looks like it would be the base right in front of the column. The easiest thing that makes the most sense is to fill the z head cavity with epoxy lead shot for mass damping right where the vibes are created. Comments?

jh

Reduce bucking, what's your mill made of!

The issue of interest is deflection due to twisting and bending, not buckling, which are both a function of the material's modulus of elasticity and the fourth power of its distance from the neutral axis. Epoxy and granite both have a low modulus of elasticity and will be positioned close to the neutral axis. If you really want increased stiffness put some additional steel structure on the outside.

If you want additional damping just fill it with oil (or water) wet sand. If your epoxy doesn't maintain intimate contact with the casting then you gain absolutely no damping.

Phil

[G59]

If you want additional damping just fill it with oil (or water) wet sand. If your epoxy doesn't maintain intimate contact with the casting then you gain absolutely no damping.

Exactly. Save yourself some money and put it towards stiffening the structure.

[R.DesJardin]

Exactly. Save yourself some money and put it towards stiffening the structure.

Or just add some hp and see what happens. "If you are going to guess, don't take all day." or in layman terms, just do it since you really can't predict the outcome. The outcome will define the structural needs if any.

[tmarks11]

Here is another alternative to the BT30 spindle you were talking about earlier:

https://www.youtube.com/watch?v=1GKG...Y&spfreload=10

Dave Decaussen ('D' in FADAL) has developed a tapered endmill holder for use in a PDB/ATC mill that uses the R8 taper, with a pull-stud drawbar that fits in an R8 cavity. Looks pretty slick. Not cheap unless you make your own toolholders (give you an excuse to buy a Tormach SPL15 lathe).

Gives you a taper with much more holding power than the 3/4" straight shank without require any modifications to the existing spindle. And I bet the drawbar could be easily retrofitted into the Tormach PDB system.

This is the same holder that he was selling earlier on the ill-fated (lamented?) Fadec UMC-10.

[johnh]

Under compression yes we are right. When you write "fixed inner structure" CMT, are you referring to reBar installed??

the derated info from, oh no, wait for it, a machining forum on the interweb!

"Oh yes. You can run the motor to 90 Hz in almost any application with only a small sacrifice in motor HP. You can run a motor satisfactorily to 120 Hz if you can tolerate a loss of about half your HP. Motor HP and torque drops rapidly from there. At 180 HZ the motor may not have enough torque to run the machine at idle.

Here'a thought for those of you in need of higher RPM and who access to a 230/460 V step-up transformer or 460 three phase. Purchase a 460 volt rated VFD. Wire the motor for 230 V. Set the V/Hz parameter to 3.83. The motor will see full hz at its full rated voltage.

As the freq setting is increased the motor runs at normal flux density to 120 Hz at full torque. If the motor is rated ar 1750 sync it can be run to 3600 and beyond. At 1 1/2 times 120 Hz (180 Hz) the motor will run to 3450 RPM - at double freq 7100 +/- RPM.

This is sound practice. There is no danger of damaging the motor armature or bearings from over-speeding within these limits."


Mine is a 60 hz motor, but if this was false info I'm all up for the full rated hp at 144hz! can I really get that Mr. wizard? ( Written with all affection looking up to those more knowledgeable than me on this subject. ) doesn't torque drop some with the increased slip at higher Hz? therefore T x rpm is not perfectly scalable?? If I want to get down to brass tac's I'll call the Sumoto tech rep. and get the skinny from the manufacturer of my VFD, on the to do list.

I totally need to see this from all perspectives so thank you sincerely!
The nay sayer comment was directed directly to any potential hard core nay sayers, not the whole list of contributors. It's an "if the shoe fits" well I'm not a shoe salesman so I'll let those people try them on for size themselves!! so attitudes will now be dismissed entirely as unnecessary input or output

Thanks CMT

jh

[johnh]

Here is another alternative to the BT30 spindle you were talking about earlier:

https://www.youtube.com/watch?v=1GKG...Y&spfreload=10

Dave Decaussen ('D' in FADAL) has developed a tapered endmill holder for use in a PDB/ATC mill that uses the R8 taper, with a pull-stud drawbar that fits in an R8 cavity. Looks pretty slick. Not cheap unless you make your own toolholders (give you an excuse to buy a Tormach SPL15 lathe).

Gives you a taper with much more holding power than the 3/4" straight shank without require any modifications to the existing spindle. And I bet the drawbar could be easily retrofitted into the Tormach PDB system.

This is the same holder that he was selling earlier on the ill-fated (lamented?) Fadec UMC-10.

Thanks for the suggestion tmarks.
As my inspection manager would say "thats funny" because we make those for him in our shop when i was there, we made a batch just a few months ago, they are tiny and the pull stud is machined solid as part of the taper, their size and the wear of a pull stud kind of turned me away from that as a consideration along with their proprietary shape.

so what do you guys think of the lead/epoxy fill right in the spindle head, nothing better than mass to absorb/dampen vibes at their source? no? harmonics aside?

adding steel to the outside doesn't make sense from the impracticality of some reasonable type of fixation at the outer most "fibers" of the castings to transfer loads.

As it was inferred by R.DesJardin, just fire it up and start stressing the thing and see what it can handle, I'm down with that since it doesn't fly!

[pickled]

That's good stuff! Makes perfect sense keen. Will definitely plan on adding that during my next "deep cleaning and PM" down time period. Thanks for sharing.

[pickled]

John I think that you have a solid plan, and I agree with not addressing any stiffness needs (developing the Tormach Exoskeleton) until you can actually gauge the level of inadequacy. Obviously FEA and modeling will only take one so far. With this type of situation the results would probably be directional at best unless you have access to a real model with all the ribbing and materials detailed. I seriously doubt that Tormach has such a thing sitting out to be downloaded seeing as they put forth a ton of work to get to the point where they have a stable design. It's not wise to give away your IP any more than you have to.