Gantry and or bit flex...

[harryball]

This is not a new issue but it has recently amplified slightly since I raised my Z axis to the higher holes. I'm wondering about how to reduce gantry flex. I'm sure some bit flex is occuring but since the problem amplified by raising the Z I'm thinking the gantry flex is the bigger issue.

I'm cutting 1/2" material with a 1/4" compression bit at 6 ips in a single pass. The required part accuracy is +- 1/32" and I'm right on the border. This last batch (first with higher Z) had some parts dead on -1/32" plus maybe just a tad. The cuts the most off are along the Y axis... i.e. the X car is stopped and the Y car is cutting with its single motor. The largest deviation is around the middle of the cut at 24". This means I'm sort of cutting bannanas.

Any suggestions are welcome.

Robert

[weslambe]

Is this a PRS? If it's not a PRS, don't read any farther. LOL

I really don't think that the extruded aluminum is flexing. That would be one hell of a lot of pressure to do that! I think the top piece is made by Bosch. Maybe not. But still, that piece isn't going to flex on a 1/4" bit since it's basically a metal torsion box.

Take a dial indicator and attach it to the z car. Place the dial indicator on the aluminum that the rails are screwed to then set a zero and move it back and forth. What this is going to tell you is if the v-rail is straight in relation to the aluminum that it's attached to.

It probably wouldn't be a bad idea to check for straightness across the top of the rails either. Since the gantry is aluminum you can simply slide your dial indicator across the top with the dial point zeroed to a spot on the aluminum bar that the vrail is attached to.

[harryball]

Sorry, it is a PRT Alpha, the flat bolt together gantry.

[conceptmachine]

Robert,
Have you checked the Z axis v rollers to see if they are snug?I had the same problem as you a few weeks ago and they were a little loose,under a heavy load my z would flex.
-shawn

[harryball]

You got me there... I have not checked the actual Z v rollers. I'll do that. I'm also adding some gusset braces that SB sent to me. They should help brace the Z tower. My next option is to slow my cutting speed, I really don't want to do that it's cutting so well.

Robert

[ed_lang]

Robert,
Please post pictures of the gusset braces.

Do they have adjustment screws so you will be able to fine tune the Z?

take care

[harryball]

I installed the gussets, I'll get pictures and post them shortly. No fine tunning on them but you read my mind... As I was installing them I was thinking about some method of doing that.

I found 2 loose Z V rollers when the Z was down at cutting depth. Good call Shawn.

All fixed up and I'll be running some cuts in a few to see how it turned out.

Robert

[harryball]

Here is one installed. I used washers to space out the long bolts to keep the ends from rubbing the springs inside. On the lower flange I drilled where an exsiting bolt held the gantry together and reinstalled the same bolt.


3760


3761

Robert

[Brady Watson]

Those are the same brackets that used to hold the PRT Z axis to the Y car way back in the bible days


-B

[harryball]

Chris with SB set me up with them. To keep my bot able to cut my projects it looks like I'm going to have to leave the Z axis in the upper holes so I'm hoping this keeps things rigid.

My second set of panel cuts last night seem to have done a little better but I still got a little more variation than before. I numbered my panels before removing them and was going to see how much and where this morning.

Robert

[harryball]

I'm confused by the data... it now looks like it is solid in the Y direction (Y car cutting along the 48" width) but my X cuts (along the len of the table) are showing variations. This time I checked and everything is tightly engaged. None of the v rollers can be turned without significant pressure with the fingers on the Z or Y trolleys. The X car also seems solid.

hmmm....

Robert

[Brady Watson]

Robert,
The little bit of flex that was occuring before was acting like a shock absorber. Now that it is really stiff, vibrations are more easily transferred to the rest of the machine.

Rather than checking the Y-car rollers, check the surfaces they ride on. If they are not smooth (rails) then hit them with steel wool or a scotch pad. Also check your pinions for wear.

Also, the dust foot mount that you made on the front of the spindle is putting a lot of strain on the t-rail. The factory mount acts like a counter balancer being mounted on the back of the rail. It wouldn't hurt to put a piece of 3/4" AL flat bar on the face of the t-rail to stiffen it up. Bolt your spindle mounting plate to it 1st, then bolt it to the t-rail. Since your Bot will cut AL, there's no reason you can't bang one of these out on the machine and try it out.

-B

[harryball]

After another test cut I am cutting bannanas along X but no longer on Y. My Y variation is minute so the Z V rollers were the likely issue. The X arc is smooth showing the greatest deflection (pert near 1/32") about mid table. I pulled some previous parts I'd cut and checked dims and it seems I had about 1/64" of variation before. I have to attribute the increased deflection to raising the Z.

Thinking about the forces at play it would seem the side force from the bit torquing which peaks at the fastest speed is somehow twisting over the Z axis or trying to lift the Y car.

Brady-
I've checked the V rails and polished them with steelwool. I didn't feel anything too harsh but there were a couple of places that felt rough. I applied some white grease to the V rails and wiped them leaving a fine film.

I checked the pinions, they still look OK. I only process about 20 sheets a month on a busy month so it's not seeing wear and tear like a daily cabinet shop would.

I don't guess I understand the dust foot putting a lot of strain on the T rail. I don't believe there is enough weight to seriously strain anything even with the added leverage. My dust hose is pulled up and secured carrying itself, when I remove the dust foot from the mount it actually hangs there and I have to pull it down and over.

As for the 3/4" AL flat bar... I not sure I follow. You mean just facing the T rail with a plate of AL and moving my spindle another 3/4" further away from my Y motor?

Since my actions cured the Y deflection do you think the added bar would reduce deflection along X which is perpendicular to the installed pate?

Robert

[Brady Watson]

Robert,
Yes - face the t-rail with a thick piece of AL. It will make it much stiffer and reduce tortion, deflection and rail wear from these forces. It would space the spindle another 3/4" further from the Y motor, which is what you want to use that weight to keep the Y car planted on the non-motor side. Deflection is less likely to occur when cutting in the Y direction than in the X. Try bending a piece of 4 X 3/16" flat bar steel into and eyebrow shape...It's easy to bend in one direction, but not the other. Same thing here. For the heck of it, put the spindle all the way down to the spoilboard + .5" and try to move the t-rail in the Y dir. Then try it in the X. You can move it in the X if you push hard enough. Now you have to imagine that you have a big gyroscope pulling the spindle all over the place, plus the added force the cutting tool exerts....

OR...Just put the Z back down a few holes and be done with it, until you really need to move it back up again. You can't expect to cut as aggressively with the Z up that high as you could with it down. You are going to get deflection unless you compensate for it by beefing things up, or reducing cutting force by not being so aggressive with your toolpaths.

-B

[harryball]

I'll take a good look at the machine and try to add some stiffness.

Moving the Z back down doesn't look like an option. I won't have enough room to cut my sides for production. The cedar sides I've been cutting are now part of my bat house production. I was cutting them manually but I prefer to let the bot do it now the hold down is solved.

Besides... I may be making this sound worse that it really is, the parts are still useable and within spec... just not as close to perfect as before. I realize when I start cutting some other jobs I may have to cut less aggressive than in the past.

Robert

[harryball]

I found something guys... don't know if I like it, don't know what to do about it but I believe I found it...

With the power on the Y car can be moved slightly. I checked the pinion and it's good and it's tight. If I pull the Y car toward Y=48 I can't budge it... if I pull the Y car toward Y=0 it shifts about 1/32". I've looked VERY close and had an observer watch... it LOOKS like the tensioning spring is giving a little and the Y car is moving but the Y motor is not moving.

Before I go taking things apart does this ring any bells, what should I look for? Is it normal? The spring has a little gap between coils, how tight is too tight?

I checked my test cuts using a straight egdge and reviewed some previous parts. It makes sense as the deflection that I always see is in the negative Y direction.

Robert

[Brady Watson]

Robert,
Not sure if it is 'normal', but I have seen this on many 1:1 Alphas before. The proper tension for the turnbuckle & spring is 3.5 turns.

You have to remember that with the 1:1 motors you are running a 20 tooth pinion. There are only 1 or 2 teeth engaged at any one time, which is why they wear much quicker than a 25 or 30T pinion. Pinions on that machine wear out way before you think they do.

If you don't like the spring setup, make some PRS style brackets that don't have the spring on them. Just make sure that A) the plates are thick enough and B) that there is enough room on the shaft to properly lock the pinion down.

-B

[rhfurniture]

Robert,
One of the best upgrades I did on my prt benchtop was to put another stepper motor on the other side of the gantry. Stopped a lot of this sort of stuff.

R.

[harryball]

Ok... so it is "normal" to the extent it is not not unusual for a 1:1 Alpha.

Short of installing a second Y motor and/or swapping out the brackets...

I should immediately change the pinion even if it "looks" good and make sure the tension is 3.5 turns.

Then I should just buck under and pay the bill to upgrade to the 7.2 motors which will improve the performance and perhaps reduce this problem.

That sound about right?

Robert

[Brady Watson]

Jessum...