A tonnage module trend monitoring

[unsaint33]

Our punch press has a SLC500 with HELM weight module (HM-604-WM) for tonnage monitoring. The operator enters the weight range on the HMI, which is moved in the ladder program to the module's output word (MOV N48:11 to O:8.3, for example). The weight module does all the calculation internally and if the weight goes out of the range, it turns on an input bit (I:8/1, for example), which the ladder uses to trigger an alarm and stops the press. I can understand this part. However, the module also monitors a "trend." It takes a set trend range (just like the regular weight range). The trend is usually set between 0 and 15%. We never change the trend range. Nobody in our plant knows ever getting a trend alarm. What is this "trend?" What is the module actually monitoring when it monitors the trend?

 

[MaxK]

derivative or FOL co-function (T*s+1)?

 

[unsaint33]

I don't know what that means. If an over trend alarm triggers, what could be wrong and what should I look for in troubleshooting?

 

[kekrahulik]

Helm is in Toledo, Ohio. Just give them a quick call.

 

[plvlce]

He is suggesting that the trend looks at not just the value of the weight at a given moment, but at what how it is changing or what it has been at over the past [time period], applying some logic to determine if something may be wrong.

For example, if the measured weight spikes back and forth between two separate in-range values without registering values in between, this would generally indicate something is wrong, but your normal alarm would not go off because the values were all inside your programmed range. Depending on exactly what it does a trend alarm might trip because this behavior shouldn't be possible. (This is a potential example only, I have actual knowledge of this module)

Unfortunately unless someone has prior knowledge of how this module functions you're unlikely to get very specific help here; the manual I could find hardly even mentions the trend.

Generally, if you get a trend alarm I would look to see whether the incoming signal from the strain gauge matches what is expected.

 

[unsaint33]

He is suggesting that the trend looks at not just the value of the weight at a given moment, but at what how it is changing or what it has been at over the past [time period], applying some logic to determine if something may be wrong.

For example, if the measured weight spikes back and forth between two separate in-range values without registering values in between, this would generally indicate something is wrong, but your normal alarm would not go off because the values were all inside your programmed range. Depending on exactly what it does a trend alarm might trip because this behavior shouldn't be possible. (This is a potential example only, I have actual knowledge of this module)

Unfortunately unless someone has prior knowledge of how this module functions you're unlikely to get very specific help here; the manual I could find hardly even mentions the trend.

Generally, if you get a trend alarm I would look to see whether the incoming signal from the strain gauge matches what is expected.

Do you happen to know the set time period? I got the manual but I haven't found any info about the trend time period setting or how to even change it.

I called Helm and had to remind their tech support that they used to make the SLC500 version before. Needless to say, I didn't get much help. So what you gave me is very valuable. Our SLC500 has two of these cards so If we ever get the trend alarm, I will swap out the cards and see if the alarm follows the card. Thank you so much.

 

[MikeyN]

I don't think the trend alarms are specific to the SLC 500 module. The trend, as you have already noted, is taken over a set time point, usually 10 strokes. During these 10 strokes, the Helm module is tracking the tonnage throughout the entire stroke. This is then averaged over all 10 strokes to set the baseline. Any deviation in tonnage above 15% will cause the trend fault. This is not a set tonnage, but is dynamic throughout the stroke. In our case, whenever ours are set to monitor, the Helm is set to start taking the 10 part sample.

Think of it like this, as simple as I can think of describing it. While the press strokes through, the Helm module is charting out the tonnage for every degree from 0 through 359. This makes a pretty nice little waveform that shows positive and negative tonnage throughout the entire stroke. Lets assume that you see the following tonnages throughout a stroke.

Fault tonnage - 250 tons
45 degrees - 80 tons (down stroke)
90 degrees - 90 tons (down stroke, meeting material)
135 degrees - 140 tons (down stroke, starting to push cams)
180 degrees - 180 tons (bottom of stroke, material is cut)
225 degrees - 60 tons (up stroke, cams pushing back on ram)
270 degrees - 80 tons (up stroke, counter balance assists)
315 degrees - 80 tons (up stroke, counter balance assists)
0 degrees - 0 tons (top of stroke)

After the Helm has created its trend, it monitors all of those tonnages compared to the angle of the press. If it sees anything above a 15% deviation throughout the stroke, it will stop the press. A scenario that we have actually seen, let's say that at 180 degrees the press sees 208 tons. This is below our fault tonnage of 250 tons, but above the 15% trend allowance. The operator noticed that the date stamp was a little light on the part and lowered the shut height, causing a higher than expected tonnage. It wasn't enough to cause an issue with the press, but the Helm module saw a problem and let us know. Since the die was already at the .050 distance allowed on the stop blocks, we had to raise the shut height back up and shim down the date stamp. In this fashion, the Helm is protecting the die rather than the press.

These numbers are all pulled from my head, don't mean anything, and are meant as a very coarse overview of how I know ours work. There are many things the trend can be used for with troubleshooting, but in my experience they are mostly application specific. Our 2000 ton draw press monitors reverse tonnage on the upstroke, and a low trend can be used to let us know the nitrogen cylinders are low and that we may have a leak.

 

[James Mcquade]

Our presses were set up for trends as well.
these trends were an indication of the load on the press due to the die, load on the press due to incoming punch load vs. outgoing punch load, bearings wearing out and other items. we were tied into these trends in the office. it was also used for individual cycles to adjust the shut height.
james

 

[unsaint33]

MikeyN, thank you so much for the great education. It makes a lot of sense. After reading your reply, I looked at our press HMI and found that there was a numeric entry button that allows us to enter the trend sample size. It's currently set for 5. I had no idea what that was. If you don't mind, here are some more questions...
1. On the same HMI screen, there was a numeric entry button for "low alarm inhibit." It's currently set for 2. What does that do?

2. Now I knkow what the high trend alarm means. But why would I need a low trend alarm? If the range is set for 0 to 15%, it means monitoring the deviation over 15% + -, right? So, what would cause a low trend alarm?

3. It sounds like I'm supposed to reset the trend whenever we change a die in order to get a new sample block. Is that correct?

4. In one press we have, the live trend reading of the right rear is consistently around 13% whereas others are only around 2%. Should I be concerned about that?

Again, thank you so much.

 

[MikeyN]

You should look at manual HM604E. This is the manual for the SLC module. you can find it here on their site.

1) This is the number of cycles a press is allowed to be low before it faults out. In your case, the press is allowed to stroke twice before it faults out to give the press a chance to get up to speed.

2) I don't believe it will be +/- 15%. The high trend alarm will be for anything 15% over what it sees as normal. The low trend can be set differently. It will be the window for tonnage below what it sees as normal. It is typical for this to be set to 0 (ignored). Perhaps taking the same shut height example, the operator didn't like a noise the press was making and decided to bump it up. The tonnage would be much lower than expected through bottom dead center, thus a lower trend fault. This is a terrible example and not typical, but the best I could think of. We do have low set because we monitor the reverse tonnage on our deep draw press. It is disabled for all of the other ones.

3) Yes, reset the trend every time you do a changeover. It is easy for us to do as our trend limits will fault the press out as soon as we unbolt the upper die. Our operators put the Helm in bypass for a changeover, and the press will set it back to monitor automatically after a set number of strokes so nobody can conveniently forget to put it back in monitor.

4) You might find that this changes after you reset the trend. If you are seeing 13% higher tonnage in that leg than the other legs, then you might have other issues. It could be that the system needs to be recalibrated, you could have an issue with the parallelism of the press, or you could have a slug under that corner of the bolster causing you to be unlevel.

Note: I assume you haven't been resetting the trend, which isn't necessarily a bad thing in this context. If you are suddenly getting over trend alarms, it does indicate that something has changed. If someone changed a shut height transducer and it wasn't calibrated a couple of thousandths low, you probably are hitting a little harder than you used to.

Helm have been pretty responsive to us in the past, and we have them come around once a year or so to recalibrate our systems for us during our winter shutdown. Maybe you should get one of their reps in to take a look at your system. They will very likely try to upsell you, but they should be able to walk over all of the settings with you and explain it better than I can. A rep looking at your system will be much more helpful than support on the phone.