CAD/CAM discussion forum > 3D CAD/CAM > QM Depreciate features. (Yet another scream of frustration!)

QM Depreciate features. (Yet another scream of frustration!)

Rank: 1

Dan

Newbie

posts: 0

Registered: 2002-8-26

Message 1 of 14

 QM Depreciate features. (Yet another scream of frustration!)
23-03-2004 02:21 . pm | View his/her posts only
This post is issued by YADFTIMAIQM (Yet Another Depreciate Feature That I Must Add In QM).

Guys QM is plagued with some depreciate features that are there only to be there because X or Y ancient CAM system that was able to mill a box (and not more than that) has it and this makes me very sad sometimes looking at it.

These techniques almost anytime generate redundant cuts and many modern machining techniques present in QM can mill these particular cases more generic and efficient.

These old and inefficient ways of milling have been there because the kind of milled parts where very simple and/or the level of mathematical apparatus present in that CAM system was empirical compared with what we have in QM these days.

I will enumerate here some of the depreciate features that you should encourage users to dont use and I explain why they are depreciated:

- Lace with variable step. This is always worse than an aggregate steep-shallow mix.

- Z Level with variable step (Thank you guys cuz u didnt push me to write it yet!) I have it but I dont want to enable it because leads to inefficient cuts);

- All finishing cycles + Automatic Equal Cusp. This technique is there only to avoid 2 steps semi-finishing or finishing but almost all the time is worst than milling with angular detection because it generates more retracts;

- All finishing cycles + Rough behavior. This is a horrible 20 years old way of milling and is working in very inefficient way only with lace or offset 2D somehow correct but still plunging in material and creating very fragmented paths. Why to use it when you have really decent roughing routines?

- Milling without AFC. This is a really serious misuse in particular for roughing and semi-finishing that can have very unexpected tool engagement during these cycles forcing a toolpath to be milled in safe mode with 30-200% longer time. AFC is a pretty nice feature adding a serious selling feature and always helps. You can use it only to understand tool engagement along the path and still is something.

- Helical motions for HSM Roughing. If is a HSM cycle in general the step in Z is small and helical arcs add plenty of motions without a real benefit a straight line motion + AFC will be more benefic.

- Trochoidal motions for HSM Roughing. This is a partial solution implemented by some CAM vendors that doesnt work always and ? of the time generates redundant cuts, generates very long NC file and keep the machine moving without a real justification. QM has wave propagation and AFC that I pretend are better technologies and more realistic in judging exotic milling scenarios.

- I dont like to rough with StepXY = [0.5, 1.00] * toolDiameter in particular in Offset2D case because very often is happen to need to add extra motions (skeletal traces) to avoid spikes (a simple loop there isnt always enough!) this require in general an extra link, approach retract etc adding many redundant motions. I recon in some cases it is justifiable BUT milling with 0.5 * ToolDiameter twice as fast or deep I make a bet will be faster! Bad things happen in Lace case too because a too big step + smoothing can leave spikes too! So a 0.4-0.45 * toolDiameter will be a decent value.

- Island Topping for Roughing. Or the today YADFTIMAIQM. This makes me proud today because I added in QM for you messing up for the benefit of some tiny particular cases efficient roughing and rest roughing only to detect prismatic flat regions, sometime detecting and modifying dynamic the Z level to synchronize Z steps to hit this flat regions. I never liked to finish flat regions with a roughing cycle. Very probable the flat region are positioning one (meeting planes) that has special tolerance requirement and milling with a variable tool loading, feeds and speeds suitable for roughing, random dilated and out of tolerance tool doesnt create (in my mad mind) good milling conditions for finishing more than this final finishing for meeting planes. So if isnt suitable for final finish why bother? Why dont you use some intermediary steps + A decent limited and set cycle for milling this special flat regions where you can explain clear the acceptable tolerance feeds and you can set a decent tool that can mill proper the flat region.

This represents my professional, subjective opinion . In general I like rich functionality and I believe QM is a statement in this regard but I hate useless aggregation and many of this not recommended ways of usage or handy functionality is a good example of mindless aggregation that can sometimes to work but very probable can be way better explained and solved using 2 or more cycles to explain the desired behavior.

And now after Ive start the storm Ill append a screen captures presenting the glorious Island Topping for Roughing.

Enjoy,
Dan

Rank: 1

Dan

Newbie

posts: 0

Registered: 2002-8-26

Message 2 of 14

 QM Depreciate features. (Yet another scream of frustration!)
23-03-2004 02:21 . pm | View his/her posts only
This post is issued by YADFTIMAIQM (Yet Another Depreciate Feature That I Must Add In QM).

Guys QM is plagued with some depreciate features that are there only to be there because X or Y ancient CAM system that was able to mill a box (and not more than that) has it and this makes me very sad sometimes looking at it.

These techniques almost anytime generate redundant cuts and many modern machining techniques present in QM can mill these particular cases more generic and efficient.

These old and inefficient ways of milling have been there because the kind of milled parts where very simple and/or the level of mathematical apparatus present in that CAM system was empirical compared with what we have in QM these days.

I will enumerate here some of the depreciate features that you should encourage users to dont use and I explain why they are depreciated:

- Lace with variable step. This is always worse than an aggregate steep-shallow mix.

- Z Level with variable step (Thank you guys cuz u didnt push me to write it yet!) I have it but I dont want to enable it because leads to inefficient cuts);

- All finishing cycles + Automatic Equal Cusp. This technique is there only to avoid 2 steps semi-finishing or finishing but almost all the time is worst than milling with angular detection because it generates more retracts;

- All finishing cycles + Rough behavior. This is a horrible 20 years old way of milling and is working in very inefficient way only with lace or offset 2D somehow correct but still plunging in material and creating very fragmented paths. Why to use it when you have really decent roughing routines?

- Milling without AFC. This is a really serious misuse in particular for roughing and semi-finishing that can have very unexpected tool engagement during these cycles forcing a toolpath to be milled in safe mode with 30-200% longer time. AFC is a pretty nice feature adding a serious selling feature and always helps. You can use it only to understand tool engagement along the path and still is something.

- Helical motions for HSM Roughing. If is a HSM cycle in general the step in Z is small and helical arcs add plenty of motions without a real benefit a straight line motion + AFC will be more benefic.

- Trochoidal motions for HSM Roughing. This is a partial solution implemented by some CAM vendors that doesnt work always and ? of the time generates redundant cuts, generates very long NC file and keep the machine moving without a real justification. QM has wave propagation and AFC that I pretend are better technologies and more realistic in judging exotic milling scenarios.

- I dont like to rough with StepXY = [0.5, 1.00] * toolDiameter in particular in Offset2D case because very often is happen to need to add extra motions (skeletal traces) to avoid spikes (a simple loop there isnt always enough!) this require in general an extra link, approach retract etc adding many redundant motions. I recon in some cases it is justifiable BUT milling with 0.5 * ToolDiameter twice as fast or deep I make a bet will be faster! Bad things happen in Lace case too because a too big step + smoothing can leave spikes too! So a 0.4-0.45 * toolDiameter will be a decent value.

- Island Topping for Roughing. Or the today YADFTIMAIQM. This makes me proud today because I added in QM for you messing up for the benefit of some tiny particular cases efficient roughing and rest roughing only to detect prismatic flat regions, sometime detecting and modifying dynamic the Z level to synchronize Z steps to hit this flat regions. I never liked to finish flat regions with a roughing cycle. Very probable the flat region are positioning one (meeting planes) that has special tolerance requirement and milling with a variable tool loading, feeds and speeds suitable for roughing, random dilated and out of tolerance tool doesnt create (in my mad mind) good milling conditions for finishing more than this final finishing for meeting planes. So if isnt suitable for final finish why bother? Why dont you use some intermediary steps + A decent limited and set cycle for milling this special flat regions where you can explain clear the acceptable tolerance feeds and you can set a decent tool that can mill proper the flat region.

This represents my professional, subjective opinion . In general I like rich functionality and I believe QM is a statement in this regard but I hate useless aggregation and many of this not recommended ways of usage or handy functionality is a good example of mindless aggregation that can sometimes to work but very probable can be way better explained and solved using 2 or more cycles to explain the desired behavior.

And now after Ive start the storm Ill append a screen captures presenting the glorious Island Topping for Roughing.

Enjoy,
Dan

Rank: 1

george

Newbie

posts: 0

Registered: 2004-3-1

Message 3 of 14

30-03-2004 06:23 . am | View his/her posts only
Why do the roughing operations in quick mill leave stock? I would like to rough out a forging die but I do not want to put in a finish pass that would add more time

Rank: 1

george

Newbie

posts: 0

Registered: 2004-3-1

Message 4 of 14

30-03-2004 06:23 . am | View his/her posts only
Why do the roughing operations in quick mill leave stock? I would like to rough out a forging die but I do not want to put in a finish pass that would add more time

Rank: 1

Dan

Newbie

posts: 0

Registered: 2002-8-26

Message 5 of 14

30-03-2004 06:38 . am | View his/her posts only
In new QM (VXV10) you can set separate thickness for frontal and radial.

QM leave stock if you ask it to leave stock in any operation. If you don't want it do not set it but be aware about the surface finish quality.

Every client that I know mills roughing with stock only to lighten the load and after this if he wants a very coarse cusp height will run a very coarse semifinish, Z Level + EqualCusp (prismatic) OR 2 operations, Z Level and area Clearance or if is shallow and sculptural Offset3D.

If you are part is already forged you can finish it or semifinish + finish according with your desired conditions. Not rough it.

In VXV10 are many new improvements in roughing.

Obvious this isn't an universal recipes.

Rank: 1

Dan

Newbie

posts: 0

Registered: 2002-8-26

Message 6 of 14

30-03-2004 06:38 . am | View his/her posts only
In new QM (VXV10) you can set separate thickness for frontal and radial.

QM leave stock if you ask it to leave stock in any operation. If you don't want it do not set it but be aware about the surface finish quality.

Every client that I know mills roughing with stock only to lighten the load and after this if he wants a very coarse cusp height will run a very coarse semifinish, Z Level + EqualCusp (prismatic) OR 2 operations, Z Level and area Clearance or if is shallow and sculptural Offset3D.

If you are part is already forged you can finish it or semifinish + finish according with your desired conditions. Not rough it.

In VXV10 are many new improvements in roughing.

Obvious this isn't an universal recipes.

Rank: 1

george

Newbie

posts: 0

Registered: 2004-3-1

Message 7 of 14

30-03-2004 06:43 . am | View his/her posts only
We are machining forging dies. On our dies we would rough it out and we would leave stock on x and y but cut z to 0. This is how we did it in our old software and it gave us very good results.

Rank: 1

george

Newbie

posts: 0

Registered: 2004-3-1

Message 8 of 14

30-03-2004 06:43 . am | View his/her posts only
We are machining forging dies. On our dies we would rough it out and we would leave stock on x and y but cut z to 0. This is how we did it in our old software and it gave us very good results.

Rank: 1

george

Newbie

posts: 0

Registered: 2004-3-1

Message 9 of 14

30-03-2004 06:48 . am | View his/her posts only
when do you predict ver 10 will be out?

Rank: 1

george

Newbie

posts: 0

Registered: 2004-3-1

Message 10 of 14

30-03-2004 06:48 . am | View his/her posts only
when do you predict ver 10 will be out?

Rank: 1

Dan

Newbie

posts: 0

Registered: 2002-8-26

Message 11 of 14

30-03-2004 06:48 . am | View his/her posts only
I told you this is how you'll do it in VX too starting with version 10 maybe quite 9.5.

You have now XY Surface Thick and Z Surface Thick.

If you have extra suggestions please post them here or submit PCRs and we will answer to them.

Usually we implement what customers wants, if is a reasonable request, relatively quick.


Rank: 1

Dan

Newbie

posts: 0

Registered: 2002-8-26

Message 12 of 14

30-03-2004 06:48 . am | View his/her posts only
I told you this is how you'll do it in VX too starting with version 10 maybe quite 9.5.

You have now XY Surface Thick and Z Surface Thick.

If you have extra suggestions please post them here or submit PCRs and we will answer to them.

Usually we implement what customers wants, if is a reasonable request, relatively quick.


Rank: 1

Dan

Newbie

posts: 0

Registered: 2002-8-26

Message 13 of 14

30-03-2004 06:49 . am | View his/her posts only
maybe 3..4 months

9.5 in less then 1.

But don't punish me if I'm wrong.

Rank: 1

Dan

Newbie

posts: 0

Registered: 2002-8-26

Message 14 of 14

30-03-2004 06:49 . am | View his/her posts only
maybe 3..4 months

9.5 in less then 1.

But don't punish me if I'm wrong.

See also