Need Help Improving Cut Quality

[whizbang]

Hi all,

So I have a Langmuir Systems Crossfire Pro. I am currently using SheetCam for camming and I am cutting with a Primeweld Cut 60.

I am having issues with bevel and corners. I eliminated the table as the source for the bevel, because the bevel is always on the same edge of the part no matter how I orientate it. I have tried to make adjustments to reduce bevel and improve my corners through some plasma cutter settings and path rules in sheetcam and while I have seen some improvements I am still not there. My corners on outer perimeters are a wreck. I should note, I am very new to sheetcam, especially using path rules.

I think part of my issue was airflow. I have a Primeweld Cut 60 that I have been using and I never calibrated the air flow to start, since doing that my bevel has reduced and actually reversed direction. My top side of the cut is narrower than the bottom now.

I attached photos of the part I have repetitively cut with various changes to settings/code. One the sides with the tabs, where the cutter is jogging in X and Y a lot in relatively short distances, my corners are rounded or not even fully cut. I used path rules to slow it down and saw some improvement, but what I can’t seem to figure out is why my inside corners on the small squares got so sharp at 60% of the feed rate, but the inside corners on the outer perimeter at 60% and 50% of the feed rate is so lousy.

Am I screwing up in my path rules? Is something to do with my consumables (even though I have used new ones)? Is it my plasma cutter?

I have a Hypertherm Powermax 85 I could wire in if someone thinks that would make a difference. At this point I need some more guidance so any and all recommendations are welcome.

[acourtjester]

There are 2 thing I would change increase the air pressure to at least 80 PSI and if using a 50 or 60 nozzle use that amount of amps. The plasma stream is shaped like a candle flame upside down so it the cutting is not at the best height you will get a bevel. I cut 1/4" steel with a 45 amp nozzle and at 48 IPM and it does very well. I would use the Hypertherm 85 at 100 PSI (be sure it does not drop below 90 PSI when cutting) at the rear of the unit and a shielded 45 amp nozzle. Stop using the path rules until you are cutting better.

[whizbang]

Thanks for the feedback.

I will adjust the air and amperage with the Cut60 and 50A nozzle. I decided to go ahead and try the Hypertherm 85, but I don't have the CPC port installed and I plan to buy a machine torch as well...so will be a couple weeks before that is ready. Any chance you can tell me what the CPC plug type is called? I don't need 25ft of cable and was going to make my own versus drop the $80 for it.

On larger parts with no path rules my cuts with 1/4" plate looked good. I did a decent size square I downloaded from this site (~ 13" x 15" with components laid out). I used my 60A nozzle cutting at 43ipm and 45A, my corners were good and the straight lines were really clean but I did have bevel issues, which made assembling the part difficult. That is when I started trying the smaller test cuts, changing out consumables, adjusting cut height, experimenting with path rules etc. I improved the bevel but made the cut worse.

Thank you again.

[weldguy]

Switching to the Hypertherm is the best thing you could have done to help yourself out

Looking at your 1st pic showing the consumables I see lot of carbon staining on the electrode and the nozzle orifice looks damaged leading me to believe your air supply is not clean and dry enough. You want your consumables to burn clean, deposits like that indicate the the plasma arc is incinerating the contamination in your air supply which will damage your consumables quickly and cause all the problems you are having.

Running that air through your Hypertherm machine and torch lead is not something you want to do. I would suggest getting yourself a better filtration and drying setup for your air supply like the Sharpe 3 stage dryer shown here https://www.plasmaspider.com/shop/dryer ... 6760.html

You will notice huge improvements right away and save money in the long run

[acourtjester]

Here are the parts needed to make your own cable, you will need to use 4 pins in the connector. They are 3 and 4 are the torch trigger input and pin 5 is the negative and 6 is the positive output from the voltage divider to the THC.
parts can be ordered from www.digikey.com

[acourtjester]

I had a guy come over to use my table to cut parts for a customer here are some of the dropouts from the 2 different size holes he had. The smaller was from a 5/8" hole not sure of the larger one. Other image is the corner of one of the small parts he cut. This is 1/4" pickled and oiled steel produced almost no dross, no cut rules used just 45 amps at 48 IPM and voltage of 127 for the THC. First image is from the back side other from top side.

[TJS]

Tom,
Got any overcut in those holes.

[acourtjester]

Hard to say it was his SheetCam plan, I don't have the plan only the G-code left on the computer. If I get a chance I will look at the start and finish position in the G-code.

[acourtjester]

Ok I took a part of the G-code from the first circle and which was about 4 lines long and single stepped through it. It stopped at the point where the lead in entered the circle, so just a complete circle with no overcut, see attached images of the cut, note the Y DRO and the X DRO for the last 2 images.

[whizbang]

So I finally got my Hypertherm PM85 with machine torch up and running with my table today. My 45A nozzles haven't shown up yet so I ran a cut with my 65A nozzle. I cut 1/4" steel at 65A, 70 psi minimum while running, and 86 ipm for the interior cuts and 75 psi for the outer perimeter. I also have a motor guard filter and water separator inline now.

My cut has certainly improved but my corners on the outer perimeter still are not square. And there is still bevel specific to the side with tabs, no matter the orientation of the part when cutting. The filter bowl on my 85 is damaged and leaks a little, a new one is on it's way. Is my issue just air pressure? Or should I also slow down my cut rates as well with the 65A nozzle?

[acourtjester]

Well you still are not high enough with air pressure, with a leak that makes it worse, and I cut 1/4" at 48 IPM. It also looks like you have a movement problem, I would stop using the path rules. There seems to be notches in the inside corners on the outside cuts this can be caused by path rules. Find a way to attach an inkpen to your torch mount and write on paper, this will help to find out if the torch is moving correctly, it also make it easier to watch the movement too.
viewtopic.php?f=3&t=29574&p=178726&hili ... en#p178726

[weldguy]

Your speed is about where it should be in my opinion, definitely want to get your air leak fixed. I doubt the leak is your issue and surprised your not seeing 0-20 or other error messages on your LCD due to the low air pressure although I am unsure if you are reading that pressure from an external gauge at the back of the PM85 or from the LCD screen.

What's your cut height at if you pause the machine mid cut on the long straight section and check with feeler gauge?

Are cutting the outside clockwise and inside cut counter clockwise?

Is the orifice in your 65 amp nozzle perfectly round and undamaged, no craters around it?

Does your 65amp shield have any of the holes plugged with slag or dust?

[TJS]

Hypertherm book speed is 45IPM with 45amp shielded nozzle. 65amp is 90IPM with shielded nozzle.
Also as others have stated air pressure into the unit should be 90psi or more. I have a gauge at the unit and when the torch fires there is considerable drop. Make sure that drop does not go below the requirement.

[whizbang]

Thanks for the input, I know the air leak needs fixing, but I am waiting on the bowl to arrive since the issue is the nut to drain water is broken off. There were no error codes, but to be honest I can't tell you what the pressure reading was when the torch ran because it was such a quick cut I didn't have time to check it.

acourtjester - Would you recommend cutting at 48 IPM even though I am using a 65A nozzle and cutting at 65A? The 45A nozzles are not scheduled to arrive til Monday. I will try to do a dry run with a pen.

I didn't use any path rules when I cut this part, I used SheetCam but deleted the all the path rules from previous cuts.

The consumables and shield are brand new, there should be no issues there since this was their first cut.

The interior cuts are cutting counter clockwise, the outer perimeter is cutting clockwise.

I have not manually measured cut height yet, but it is set to 0.060" in the program with a

[acourtjester]

I have never used a 65 amp to cut 1/4" only 45 amp shielded and at 48 IPM. With the cut height of 0.060" without a THC there is no guarantee that is maintained throughout the cut operation. As to the rules I found you needed to go to the tools drop down menu and click on cutting rules and in the box be sure the none is entered or they may still be added (les may have fixed that)

aaa rules.JPG (59.58 KiB) Viewed 2581 times

aaa rules.JPG (59.58 KiB) Viewed 2581 times

[greg9504]

he tools drop down menu and click on cutting rules and in the box be sure the none is entered or they may still be added

This just happened to me the other day, so I would say no it's not fixed yet Although I may be an update behind.
If you still have path rules applied you will see this in the displayed path. The path will be a light blue/white where the rule is applied and green elsewhere. If your path is not all green, you have rules applied.

To the original poster, attach your gcode.

Also check that your torch is square to the table, remove the lower mount so you can get a square up against the torch body. I had to resquare mine.

[whizbang]

When I am in sheetcam the path is entirely green. But I was still having issues so I went as far as to completely delete the path rules. Still to no avail I cannot get good cuts around the tabbed area.

I have corrected my air leak issues and am using a good separator and filter now so I believe I have eliminated air supply issues. I manually maxed out the air for the machine and had a mess in the shield, and to be honest, even when I let the machine regulate the air I still get ALOT of splatter up in the shield.

I have now tried 45A nozzles at 45A and 48ipm, 65A nozzles at 65A and 90ipm and 85/75 ipm and the corners are still a mess and I get varying degrees bevel in both directions and now I have managed to get both concave and convex edges, in the same cut on 2 different edges.

I have squared up the torch again and I have re-orientated the part and the problem follows hte edge so I honestly don't think it has to do with my table movement.

I have a CrossFIre Pro table with THC. I had been using Smart Voltage but finally tried a cut with both 65A and 45A where I manually set the voltage and still no improvements.

I attached the dxf file for anyone who wants to prove it is me, my machine settings etc. Site won't let me attach the .tap Gcode file from sheetcam right now.

[whizbang]

Here are some photos and details of the most recent cuts

[acourtjester]

In you file this image shows the shield with a lot of splashed up metal, What is the torch height when you are doing the pierce, it should be 0.060".
I would load your G-code (plasma off) and do a single step through it until you are at the pierce point then measure the gap between the metal and the torch shield. Not sure if your software has this option.

nozzle.JPG (24.18 KiB) Viewed 2538 times

nozzle.JPG (24.18 KiB) Viewed 2538 times

[whizbang]

acourtjester, good catch. My Pierce height was supposed to be 0.15" but I had a typo and it was set to 0.015" in the sheetcam tools. It was set to 0.15 in Fusion 360 though.

I am using Firecontrol by Langmuir and I am not sure I can step through it line by line, but I will try.

Obviously that would fix the slag build up on the shield, but my cut height for the rest of the run is 0.060". Would that really have that much of an affect the remaining cut? My straight side has always continued to look really good.

[acourtjester]

I have not started working with Fusion but I don't think the height will carry through, since you are using SheetCam for the G-code generation.
Another point is are you using a Post Processor for the Langmuir table, Do you have a way to detect the metal surface before each cut operation. Either a floating head switch or Ohmic sensor, if not you cannot be sure of the torch height when cutting.

[whizbang]

I ran one Gcode program generated by Fusion, my 3rd cut (3rd slide). The rest have been from Sheetcam.

For both Fusion and Sheetcam I am using a post processor for the Langmuir Table. The table has an IHS (Initial Height Sensing) system built into the Z axis, so I have never done anything else to detect the surface.

[acourtjester]

that sounds good if it works, I have no connection with that info, I would assume their manual tells how to set it up, and test.
Keep plugging at it and things will get working better, one thing is you are checking things and are learning. I am not the expert just trying to help.
Things generally work fine using a manufacturer's post processor in SheetCam. There could be a problem in the tool setup in SheetCam as you found earlier with the pierce height. In this video you can see how the sensing works with a Ohmic sensor, Other info in the video are for a different reason.

[whizbang]

So I did a dry run on my corrected GCode, now that I corrected the pierce height from .015" to .150", I have confirmed that heights are correct. I replaced my electrode and tips and re-ran the part, yet again, at both 45A and 65A with the corresponding nozzles.

I used the Sheetcam "Looped Corners" and greatly improved my exterior corners and my interior corners also improved with the slower speeds I used on the outer cut, but they are not quite where I need them to be. THe parts I intend to cut for purpose go together with tabs like puzzle pieces. I am looking into looping interior corners now.

[acourtjester]

There still seems to be a problem as the bottom of the inside corners look very rounded. I have no experience with a Langmuir table or their Smart Voltage operation but it seems you may not be cutting at the correct height. With that being said in the video I supplied above you can see that when testing the cut operation was stopped at a point after the THC had control to actually measure the cut height to be sure it was at 0.060".
I can tell you that a cut chart like the one Hypertherm has in their manual is only a guide. I have used different plasma units and different THC units and the only true way is to run test cuts to find the correct voltage for your table.
there have been many post about this problem in the past here is one.
viewtopic.php?f=52&t=28948&p=175329&hil ... ng#p175329