How do I stop creating lead-in divots?

[chadmjohn]

I've tried a number of settings on my table (CandCNC setup), and overall it cuts quite nicely, except for lead-in divots (not pierce divots). I've scoured the forums and tried all the things I could find / comprehend, but no success in improving the situation.

Example setup for 16g cold rolled, mild steel
40 amps
210 ipm
0 or .1sec pierce delay
anywhere from .08 - .12 pierce height
.06 cut height

FYI, my heights in the machine settings below are offset by .03". The my torch shield had a negative .03' nozzle offset, so that gets added to the settings. The .03" in the settings below is actually a .06 cut height.

The cuts come out pretty well, with minimal dross for such thin material, but I ALWAYS get a lead-in divot. I've tried arc and straight line lead-ins with no difference. I even tried setting a .2 inch overburn to see if it would "blend it", but nothing I do helps.

Here are some pics of the cut (front and back), a zoom in of the lead-in divot, and a pic of my settings (just to be thorough).

Thoughts and input would be greatly appreciated!

Front

Zoomed

Backside (fresh off the table, very little dross)

Sheetcam Settings

[acourtjester]

the first thing I would change is the pierce height to 0.160" and maybe the delay to 1.3 seconds. Your setting in the Jet tool are way off pierce height of 0.040" delay 0.0 cna cut a 0.030". Also I would drop the pause at end of cut. If you have a problem where the plasma is not on long enough to finish the cut maybe then.

[chadmjohn]

the first thing I would change is the pierce height to 0.160" and maybe the delay to 1.3 seconds. Your setting in the Jet tool are way off pierce height of 0.040" delay 0.0 cna cut a 0.030". Also I would drop the pause at end of cut. If you have a problem where the plasma is not on long enough to finish the cut maybe then.
pierce.JPG

Sorry, I forgot to add this to my original post (I just updated it)

my heights in the machine settings below are offset by .03". The my torch shield had a negative .03' nozzle offset, so that gets added to the settings. The .03" in the settings below is actually a .06 cut height.

with 16g, the pierce is nearly instant and very little topside dross gets created. I incrementally lowered the pierce height to reduce the size of the pierce hole on open lines. However, the divot you see is not related to the pierce, the pierce is .2" back on the lead in.

I hadn't thought about the pause at the end of cut causing it, it may be that it's not on the start, but on the end the divot is being created (as the arc sits there). I'll give that a shot, thanks for the suggestions!

[chadmjohn]

the first thing I would change is the pierce height to 0.160" and maybe the delay to 1.3 seconds. Your setting in the Jet tool are way off pierce height of 0.040" delay 0.0 cna cut a 0.030". Also I would drop the pause at end of cut. If you have a problem where the plasma is not on long enough to finish the cut maybe then.
pierce.JPG

Removing "pause at end of cut" did dramatically reduce the divot, but it is still there. I hit this one quick with the sander to make it a bit more clear. This is the front side and you can see how much better it is (thank you for that), but I'm still getting a divot.

Is it possible that I'm exceeding the speed of my table and the transition from the lead-in to the cut is just taking too long?

Front Side

Back Side

[FabLab]

I would recommend changing the lead-in to either arc or tangent, I normally have good luck when using an arc lead-in with a length twice that of the material thickness

[adbuch]

the first thing I would change is the pierce height to 0.160" and maybe the delay to 1.3 seconds. Your setting in the Jet tool are way off pierce height of 0.040" delay 0.0 cna cut a 0.030". Also I would drop the pause at end of cut. If you have a problem where the plasma is not on long enough to finish the cut maybe then.
pierce.JPG

Removing "pause at end of cut" did dramatically reduce the divot, but it is still there. I hit this one quick with the sander to make it a bit more clear. This is the front side and you can see how much better it is (thank you for that), but I'm still getting a divot.

Is it possible that I'm exceeding the speed of my table and the transition from the lead-in to the cut is just taking too long?

Front Side
275383548_1068530037059993_7595102288563948852_n.jpg

Back Side
275369967_476511574135850_2556019866906083773_n.jpg

This divot may be caused because you are cutting the same spot twice - once at lead-in and once at the end of the cut. I would try stopping your cut about 0.020" or so before the actual end of your current cut path. Plasmacam Design Edge (which I use) has a setting for this. I am not familiar with your control program, but perhaps it has a similar setting. Or else do it manually by modifying your cut path or the drawing you are using to create your cut path. You can experiment with some different settings to see what works best for you.
David

[weldguy]

Indeed experimenting with your lead in type or angle will likely really help. I always try to lead into the part at a very shallow angle or no angle at all. Often the sharp direction change at the end of the lead in will cause what your seeing.

[TJS]

Indeed experimenting with your lead in type or angle will likely really help. I always try to lead into the part at a very shallow angle or no angle at all. Often the sharp direction change at the end of the lead in will cause what your seeing.

Can you show where that is in sheetcam. I am not seeing that setting for changing the angle of the lead in. Great idea.

[acourtjester]

Here you go

plasma 5.JPG (94.99 KiB) Viewed 4225 times

plasma 5.JPG (94.99 KiB) Viewed 4225 times

[TJS]

Yes I have that, but where do you actually change the angle of the lead in, is what I want to know. Like changing the angle of ARC itself.

[djreiswig]

I don't believe the angle is adjustable.
Maybe post your question on the sheetcam forum.

[weldguy]

Indeed experimenting with your lead in type or angle will likely really help. I always try to lead into the part at a very shallow angle or no angle at all. Often the sharp direction change at the end of the lead in will cause what your seeing.

Can you show where that is in sheetcam. I am not seeing that setting for changing the angle of the lead in. Great idea.

I am not a Sheetcam user so I am not sure, maybe that option is not available with Sheetcam.

[SeanP]

I find perpendicular leadin best like you have there, maybe a touch longer, I would say the divot is from not having a leadout though, I usually use a arc lead out, 1/8'' will do it.

[chadmjohn]

Yes I have that, but where do you actually change the angle of the lead in, is what I want to know. Like changing the angle of ARC itself.

I use sheetcam, but I don't know if you can actually adjust the lead in angle.

The way I interpreted the suggestion was to start the lead in on a straight edge instead of an arc. In my example, the lead in starts on the outer edge of an upper case "S". It may make less of a divot if the start point were changed to the relatively straight segment between the upper and lower arcs of the "S".

[djreiswig]

I try to stay away from putting start points in the middle of straight paths. If you put them on corners or ends of letters it doesn't make them as obvious.

[adbuch]

I try to stay away from putting start points in the middle of straight paths. If you put them on corners or ends of letters it doesn't make them as obvious.

[Blackslacks]

Try setting up a cut rule to turn your torch off slightly before end of cut. As is, the torch is stationary for a few milliseconds before the arc is actually extinguished. This may be causing the divot.

[adbuch]

Try setting up a cut rule to turn your torch off slightly before end of cut. As is, the torch is stationary for a few milliseconds before the arc is actually extinguished. This may be causing the divot.

I mentioned this before, but had no response from the OP. In Design Edge, we do this by setting the "gap at and of loop". You have suggested how to accomplish the same result with Sheetcam settings (rules). Perhaps the OP will try this and let us know if it has helped. I do know of other instances where this technique has been use to great success.
David

[TJS]

Bringing this back to the top. Did the OP get this minimized even more. I am cutting 18ga and I am getting a divot too. It is my inside offset cut. I just came across this after I searched "divot". I will try removing pause at end of cut. I will also try Perpendicular lead in and also arc a lead out. Not sure I want to stack all these changes up at once. I do have 25 parts to cut out but want to get this right for the customer because he pointed out the test sample I gave him.
T.J.
I am running Command CNC from CandCNC and they state do not add cut rules because it messes up the canned cut rule at "before end of cut" cut rule.

[N2 Welding]

Adding a question I have for learning purposes. OP has cut feed rate of 210ipm when I do the math that comes out to 3.5 ips. DTHC Delay -sec = 1

I was wondering if the delay in dthc of 1 sec could cause a problem as I don't know if the torch actually moves to cut height before it starts moving after the pierce. Then after thinking about it, if that was the cause then about 3.25 inches would be effected and not just the lead in/out affected zone.

Be interesting to see if the end of cut is the cause. I’d think stopping the cut about an inch into cutting should be enough to show if it is coming from the start. Then if it does not look like it is coming from the start of the cut shift the focus to the lead out and end of cut settings. Does this sound about right ???

The way I interpreted the suggestion was to start the lead in on a straight edge instead of an arc. In my example, the lead in starts on the outer edge of an upper case "S". It may make less of a divot if the start point were changed to the relatively straight segment between the upper and lower arcs of the "S".

I think I read or saw a video that talked about lead in’s / out’s should be done on the straight edges instead of at curved edges or maybe it was the other way around. I forget already. However that would not fix the issue if one is cutting circles etc where no straight edges exist and on the flip side if one is cutting only straight edges.

[acourtjester]

To make the torch action clear, there is a pierce delay and a THC delay. The pierce delay is to be sure the metal is pierced through before moving the torch. The THC delay is to allow the arc to stabilize before the THC starts functioning after the torch is moving. The pierce delay is short like 0.5 seconds, the THC delay is like 1 to 1.5 seconds. The term DTHC is mostly used by CandCNC users (Digital Torch Height Control) same as THC by others.
You can select the path of the lead in/outs as Arc, Tangent, Perpendicular or None, placement is a learned choice as to their affect.

[N2 Welding]

So does the torch go to cut height after making a pierce and before it starts moving x & y axes or does it start moving the torch x & y axes before adjusting torch z axes for cut height?

[robertspark]

torch touches off and raises to pierce height
torch fires
torch waits for transfer signal (if THC enabled!!!)
once transfer signal received (if THC enabled)
pierce delay timer starts (0.5 seconds or whatever you program) AND (if THC enabled AND you have the THC delay, the THC delay timer starts)
at end of pierce delay
torch either drops to cut height OR (if you have AND need the function) raises to puddle jump height
torch moves in XY direction.

IF puddle jump being used, after set distance, torch drops to cut height

after end of THC delay time period (IF enabled or available), THC becomes active and Z (up and down motion) to track the voltage begins.

if the voltage is LOW (below setpoint) the torch RISES if the voltage is HIGH (above setpoint) the torch LOWERS.

hence on small arcs, holes or corners turn THC off as the torch will lower as the voltage rises because the kerf will get wider.


some THCs don't have THC delay (time) and some THCs don't have puddle jump (delay), and it's only needed really on thick material where there is a lot of blow back at time of pierce and the molten metal forms a "volcano" shaped crater (lots of top side dross) so that the torch does not crash into or plough through the molten metal (damaging your tip and possibly giving a bad voltage reading if there is slag on the tip)

[N2 Welding]

Very good explanation Robert thank you

[robertspark]

no problem

here is a really good document, some of it you may know, but some of it you may not... just read it all take your time

http://linuxcnc.org/docs/devel/html/pla ... rimer.html


this is another document, yes it is specifically for qtplasmac, just ignore that for a moment. skim through it to understand the functions and features that a fully fledged plasma cutter CNC controller/ machine should have. then compare it to whatever you currently run. then you can appreciate what you have and what is available.

linuxcnc is not easy, there is a lot of lingo behind it, and it does have a steep learning curve when you want to install something and it does not just install. yes I know that Debian will soon fully include linuxcnc

http://linuxcnc.org/docs/devel/html/pla ... asmac.html

sure it may be difficult, but so was learning to walk, read, write, etc.....