Photo-Eye Blocked or Not?

I am simulating two container filler lines to display the buffering capabilities of one of our pieces of equipment. Each line is using photo-eyes to control product consumption rates of the downstream processor. Occasionally, especially while using the experimenter to iterate results, one of the lines stops. Looking at the model during one of these events shows a photo-eye that displays uncovered while clearly being covered. I have attached captures of what I would consider normal (OK), and abnormal (not OK) photo-eye conditions.

Has anyone else experienced this?

I will move the photo-eyes marginally to see if that helps, but the placement is important, as they are located where the customer has specified.

Thanks,
Peter

Is flow going left to right in the captured images?

Flow is right to left in the captures. Essentially I am using the photoeyes as prime and hi/low switch to modulate processor speed. If we lose the prime eye (closest to processor), we stop the processor. If we lose the pace eye, we set slow speed; blocked pace eye sets high processor speed.

I've never seen this, but maybe someone else has. If you can reliably reproduce this problem, would you mind posting the model with a set of instructions on how I can go about reproducing the problem for myself? Thanks.

Cliff,
I will focus on replicating the error today and post the project.

Thanks
Peter

I am still trying to reliably reproduce the error I showed in the screen captures, but came across an error just as disturbing with conveyor photo eyes:

Set stop time to 205, run. Fillers go through MTTR event.

At 207, Conveyor3 and DynacPacker finish a MTTR event. Flowitems on Conveyor3 'jump' to a fully packed condition. Reservoir360 releases flowitems, fully populating conveyor3.
Photoeyes on Conveyor3 remain inert and inactive until approx time 233, then register product transition/blockage.

This error seems to occur when a changeconveyorspeed command is issued while the conveyor object is stopped.
To replicate:
Set stop time 630 (get past all the MTBF/MTTR stuff to steady state)
Open GUI - click Stop Fillers button.
When Reservoir <5% full, click Stop Conv3 button.
Click Start Fillers button.
When Reservoir >5% full, click Start Conv3 button.

I have also removed my changeconveyorspeed commands (Reservoir triggers rise through middle and fall through middle) and done this test without the error occurring.

v4.3_2

Thanks
Peter

Peter,
Every time I try and follow the instructions that you posted, Flexsim freezes (becomes unresponsive) which is usually evidence of an infinite while loop. Does this happen to you as well? I will keep running tests and checking for this, but that may be something you will want to look into.

Dustin Derrick
Flexsim Support

Dustin,
I was able to run it normally and with experimenter without issue. I did not use any while loops in the scripting, so I'm not sure what would cause that. I will download the version I posted and verify operation against the saved version on the PC.

edit: I edited the MTBF start time to make it stay where I wanted - apparently, editing the value in the GUI isn't saved through a close/open. I am also including the wrl files used (place in fs3d directory - all are from this site except Dynac6400v2, which was created by Bill N.). The project still ran without error (at least not freezing, anyway).

Thanks
Peter

Peter,
To make sure I understand what the problem is, is it that the " Photoeyes on Conveyor3 remain inert and inactive until approx time 233, then register product transition/blockage." or is the problem that "Flowitems on Conveyor3 'jump' to a fully packed condition." Or is this the same problem that you mention in your first post? I think I am a little confused as to what the problem is and what you want to see happen. If you could clarify this for me a little it would help.

Thanks,

Dustin Derrick
Flexsim Support

Dustin,
In attempting the recreate the first problem, I came across the other two:

" Photoeyes on Conveyor3 remain inert and inactive until approx time 233, then register product transition/blockage."

"Flowitems on Conveyor3 'jump' to a fully packed condition."

So, I guess I have three problems total.

Since my model has changed a good bit since the first issue, I may have invariably created a work-around. I had been asked to move on to complete the model, so the troubleshooting had to wait. As I recall, I added some hysteresis on the Reservoir to delay releasing parts for several seconds. I will go through the older versions this morning to find the initial issue.

Thanks,
Peter

Peter,

I finally got an opportunity to look at your "photoeye issue.fsm" model, and after poking around, I have the following suggestions for you. These suggestions are important for all your models involving conveyors, MTBF/MTTR's, photoeyes, and especially the use of the changeconveyorspeed command.

1. There's a warning in the help regarding the opening of ports (i.e. openinput) on an entry or exit trigger. The same could apply to the cover and uncover events of photoeyes. It's not a guaranteed problem, but circumstances could converge that would make it a problem. It's just a good idea to not use open or resume commands during a transition of the object. The way around it is to use senddelayedmessage() with a delay time of 0 is ok, and then do the opening or resuming in the OnMessage trigger.
2. Do not execute the changeconveyorspeed command on a conveyor in a down state. This you do in the rise/fall through middle mark of Reservoir360. Maybe use a conditional statement to check the state of the conveyor first.
3. Try not to use a combination of stopobject/resumeobject to control the state of the conveyor and changeconveyorspeed to manipulate the conveyor's speed. I suggest you use changeconveyorspeed to control the start and stop of the conveyor as well as for changing its speed. You can easily bring a conveyor to a stop by changing its speed to 0 in the OnStop trigger of the MTBF/MTTR, and then in the OnResume trigger set it back to what you want its resume speed to be. You will want to remove the conveyor from the member list of the MTBF/MTTR if you use this approach.

I hope these suggestions will resolve the problem with this model and maybe some of the "weirdnesses" you've been experiencing with your other models.

Good luck as you work through these more complex issues. Photoeyes and dynamic speed changes to a conveyor are pretty advanced features. Your model looks very cool by the way!

Thanks for the pointers, Cliff - I'm still deep in the learning curve. One of these months I'll have to sign up for training. The model is representative of our business, conveyor systems, which happen to be laden with photo eyes and dynamic speed control. I'll definitely revise with your suggestions.

Thanks,
Peter - looking forward to the bottling library