What did I do wrong (Part 2)

Well,I have cut my first gear. Thanks to all who contributed,but, and there is always a but. The circumference is identical to the pattern.The thickness is identical to the pattern. The No of teeth cut is correct,the depth of each cut is identical,but.....some (3) of the teeth are thicker than the others and 2 are thinner. 11 are perfect. What could I have done to cause this ?

Not moving the sector plates fully? Not going in the same direction so backlash is introduced?

Because I was learning how to do this,I did a dummy run and found several points that needed attention. The two that you mentioned were covered,together with basics like checking the tightness of holding down clamps after each cut. Making sure there was enough clearance between the cutter and the chuck jaws when the cutter was returned for a fresh cut. I thought I had done enough to ensure near perfection,but I was wrong,I obviously missed something.

How are you holding the gear blank did it move?

Did you lock the table for each cut?

Edited By JasonB on 05/11/2015 15:43:47

Sounds like backlash to me. Make sure it's taken up before each cut and in the same direction.

Martin

As the depth is correct on all it sounds like a back lash problem to me or the index fingers loosing their spacing if really bad.

I did look at various rotary thingies at the recent model engineering exhibition and wasn't impressed with degree of backlash on the ones that could be tried. In fact the adjustments were limited so that it was impossible to check if there were tight spots and how bad these might be.

I'm not at all sure how some one could check that the back lash is still taken up when the clamp is released ready to move to the next index so over to some one else.

John

-

It is possible that the blank moved on its arbour, it needs to be really tightly held. Colin

The gear blank did not move as it was Loctited on to a tapered mandrel with the direction of the cut pushing it further on,which it couldn't as I had a spacer, and I had a very difficult job of getting the gear off ,afterward (hot water and repeated hide hammer blows) However....locking the table,I did not do,nor do I see how this can be done. I was not aware of any lateral movement whilst I was cutting, but I could be wrong as there appears to be no way of locking this. If there was any linear movement this should not make any difference,should it ? I am beginning to think it could be lateral movement,as that would explain the difference in teeth width,but I don't see how to cure it.

Hmm when I cut my first bevel gear something similar happened, a couple of teeth were clearly off. I was using a rotary table with an indexing plate. Turned out the chart that came with the indexing plate had a couple of duff numbers in it - The problem only became evident as I got to the last teeth.

Steve

Did you ever overshoot and have to wind the head back?

If so you should always wind back by a good half turn.

Neil

What table is it as most have some form of lock

It is a 1956 Pools horizontal milling machine. Very nice and easy to use.

I have just reread the Lathes.com article on this machine and it says "the x and y axis can be locked" !!!!!!

Investigation tomorrow.

Roger

Rodger you state that gear blank was secure on the mandrel,how is the mandrel attached to the rotary table? if the workpiece is slipping it usually just slips in one direction so all all the errors would be say thicker teeth, except for the last one which would be thin, slippage will occur if the edges of the teeth on one side of the cutter are blunt compared with the other side,the blunt side tends to push the workpiece sideways.The error is also possibly due to poor indexing, when indexing the next division the plunger index pin is carefully pulled from the hole in the plate and usual practice is to rotate the index arm and pin clockwise,move the arm carefully until the pin is nearly inline with the correct hole in the index plate,then ever so carefully just tap the arm with your knuckle until the pin drops in the correct hole,then lock the rotary table, if the arm and pin goes too far and misses the correct hole,turn the arm anticlockwise for a quarter of a turn then try again,this will eliminate backlash problems.Also vibration and chatter can move a workpiece,on larger teeth it is sometimes better to cut at 2/3 depth all the teeth and then go round again to finish depth, some operators prefer to cut to full depth to avoid rubbing of the cutter on small cuts, but it is better to get a good gear after going around twice, rather than try to cut to full depth and ruin the work.

It was the rotation of the table I was suggesting should be locked but the x & y of the mill should also be locked

I have discovered how to lock the table,it was quite simple really.

Again,thanks for these advices (is that good grammar ?) I had not been locking the x and y axis nor had I been locking the chuck prior to each cut.

I am now going to embark on cutting my next gear,hopefully following all this advice will make my effort more acceptable.

Roger

A rotary table can normally be adjusted for backlash, it should have the worm assembly mounted in an eccentric bush.

You can use it to set the depth of engagement with the worm wheel to get rid of backlash, it is also used when you want to disconnect the worm completely so you can spin the table by hand for setting work true etc.

Phil

That's what I expected Phil but having looked at a number of them at the model engineering exhibition none would allow it to be fully taken out. Some sort of stop inside. Suggests that Neil's wind back 1/2 turn and then to the sector arm is a must.

There was model engineer design about that use a spring to tension the worm against the wheel. It also uses worm pressure via a cam to lock it. Not too keen on that but the spring idea seems sound.

John

-

Bit of string wound around the table with a weight on the end would ensure the backlash is taken up if you are thinking simple fixes. Best to do it so the index arm raises the weight.

Martin