"Synchronising" product flow on two conveyors

[markymark31]

Hi All,

I am at the design stage of this one and looking for some pointers please ? Has anyone done anything to "synchronise" product flow on two parallel conveyors ?

Basically I have got one conveyor that runs at a fixed speed and it has a number of "trays" fixed to it.

I then have a second conveyor that carries my product that I have to "drop" on to each "tray".

Using PEC's, encoders, and shifting the memory in the PLC I can "track" the tray position along the first conveyor, and also the product along the second conveyor, this part is not a problem.

What I then need to do is work out how much I need to speed up / slow down the second conveyor to ensure that the product drops on to the correct position on the "tray".

The second conveyor is broken into three separate drives at the "outfeed" end, each controlled by a VFD so I can adjust the speed as required.

What I am thinking of doing is checking the leading edge of the tray at a predetermined point before the "drop" point and working out how far away the leading edge of the product is, then I can adjust the speed of the second conveyor to suit.

Has anyone done anything like this before ? can you give me any tips ?

Thank you,

Mark

 

[frankh]

Synchronization Controller Needed

Hi,

It looks like you are in need of a synchronization controller.

Lenord+Bauer offers this proven technology. Take a look here:

http://www.lenord.de/en/produkte/motionline/EcoController/gel8140/db8140_e.pdf

And also here: http://www.lenord.de

Kind regards, and good luck!

 

[MichaelG]

I think that I have done something very similar.

What PLC platform are you working on?

Are you are essentially trying to match phase on the first conveyor with random placement of product on the second conveyor using 3 short belts?

Some people call them smart belts.

 

[markymark31]

Hi frankh thanks for the link, I will look at this product.

Hi MichaelG I am using an AB CompactLogix PLC. I am not sure about your term "match phase" but I think it is essentially a "yes" as I am looking to use the three short "smart" belts to adjust the position of my product to match in / sync with certain positions on the main conveyor.

Mark.

 

[Peter Nachtwey]

Once geared one to one, the product will be dropped on the fixed speed conveyor at the same place all the time. If you are lucky the product will be dropped in the middle of the tray but if the product is dropped on the leading edge of the tray you may need to adjust the phase of the second conveyor by delaying it by a few counts. This is phasing. Matching phase allows you to drop the products in the middle of the tray. Not one the leading or trailing edge of the tray or even outside of the tray.

Phasing is usually done by superimposing a move on top of the geared move. In our case we superimpose a half sine or a 5th order motion profile. In the command you specify how far + or - the slave ( the second conveyor ) needs to be offset from its current position. Another parameter specifies how far the master will travel while the slave is changing its offset.

You can do this with a RMC70 with one axis and one reference encoder input for the fixed line encoder. The RMC70 has gearing and phasing commands that would make synchronizing two conveyors trivial. The RMC70 communicates with the Compact Logix using EtherNet/IP.

This is a much more difficult application with lots of synchronizing and phasing. The second part of the video shows the leading edge of the plywood sheets blocking photo eyes. The motion controller must even up the leading edges by speeding up or slowing down one side of the conveyor to the edge of the plywood lines up with the laser line. This phasing must happen on-the-fly. This was all done in the motion controller except for moving the big billet of layered plywood below the gates. That motion wasn't fast or critical and was done by a PLC5.
http://www.deltamotion.com/peter/Videos/Auto Lay Up.wmv

Another good example is cutting logs. When the saws cut the logs to length there is only a little space between the two pieces of wood. To increase this space so the longs can be dropped horizontally the down stream chain makes a small super imposed ( phasing ) move on top of the geared or synchronized 1 to 1 move that increases the distance between the logs.

 

[OkiePC]

What Peter says first is critical. What controls the "dropping" of product onto the first conveyor? Can you control that part of the operation? It needs to be matched to the 2nd conveyor tray pitch frequency. Then the conveyor in the middle can run at a follower speed, but speed control needs to be tight throughout the speed range of the tray conveyor...since you say it's fixed, then the first conveyor can be fixed, with some adjustement and should do pretty good, but only if you control, or can accurately detect a very repeatable product entry frequency.

If you control the entry of product onto the first conveyor, then you can simply gear them 1:1 as Peter says, and have a means to offset the phase to compensate for timing errors (if any) from sensors and actuators.

You at least will need full vector VFDs with tight speed control, but a servo system would make this easy, and could be necessary if you can't control the product entry timing. If it's variable, and the tray conveyor is fixed, then your variable conveyor needs to be a smart indexer...which can be very difficult if you can't at least interlock with the upstream producer. (Tell it to stop, if your indexer is required to decelerating to match the next tray.)

Does the initial product entry occur at a known, predictable, stable rate, or is it pretty random?

 

[Peter Nachtwey]

What Peter says first is critical. What controls the "dropping" of product onto the first conveyor?

In markymark31's application I am assuming that the widgets drop onto the fixed conveyor. However, the controller can activate a digital output to trigger the dropping but this shouldn't be necessary.

Can you control that part of the operation?

Easy.

It needs to be matched to the 2nd conveyor tray pitch frequency.

Again, this is easy.

Then the conveyor in the middle can run at a follower speed, but speed control needs to be tight throughout the speed range of the tray conveyor...since you say it's fixed, then the first conveyor can be fixed, with some adjustement and should do pretty good, but only if you control, or can accurately detect a very repeatable product entry frequency.

It is extremely easy if the product on variable speed conveyors are equally spaced. Once phased then nothing more needs to happen. If the widgets are not equally spaced we can still handle that. Obviously the phasing will need to change for each widget but that isn't any different from the video where the phasing had to be done for each sheet to align it with the laser line.

If you control the entry of product onto the first conveyor, then you can simply gear them 1:1 as Peter says, and have a means to offset the phase to compensate for timing errors (if any) from sensors and actuators.

Yes, the job becomes much easier if the widgets are spaced evenly but this isn't a requirement with a good motion controller.

You at least will need full vector VFDs with tight speed control, but a servo system would make this easy, and could be necessary if you can't control the product entry timing.

I think a VFD with good speed control would do. It depends on how precise the dropping of products must be. Dropping a widget on a tray doesn't sound like it requires a lot of precision. So far precision doesn't sound like the problem. It is the algorithms that are a problem unless they are already programmed in a motion controller..

If it's variable, and the tray conveyor is fixed, then your variable conveyor needs to be a smart indexer...which can be very difficult if you can't at least interlock with the upstream producer. (Tell it to stop, if your indexer is required to decelerating to match the next tray.)

Like I said, this is easy. There are some gotchas though. If the spacing between the trays and the spacing between the widgets varies a lot there will be problems. For instance. If the trays are 1 meter apart and for some reason the space between two widgets if 3 meters apart and the variable speed conveyor will need to travel 3 meters while the fixed speed conveyor travels 1 meter. This would mean the variable speed conveyor may need to move very fast and this may not be desirable. In this case it may be wiser to skip a tray so the variable speed conveyor travels 3 meters while the fixed conveyor travels two meters.
These calculations can be done in the motion controller too.

Does the initial product entry occur at a known, predictable, stable rate, or is it pretty random?

Good question but it doesn't affect the degree of difficulty much if using a good motion controller.

We have a really slick command called the Advanced Gear Move. You specify the slave's position, gear ratio and gear ratio rate when the master reaches a position. In this case the master is moving 1 meter between trays. If a photo eye is detecting when the widgets are entering on the variable conveyor the motion controller can compute the optimal gear ratio and gear ratio rate that would make for a smooth transition to the next widget. A good motion controller can keep track of a FIFO of widget spacings or locations. For instance if the widgets are 1/3 of a meter apart then the gear ratio would be 1 to 3 but what if the next part is 1/2 meter apart? Then the gear ratio would change to about 1 to 2. You can see that if the widgets are not evenly spaced then it would be good to compute the optimal gear ratios on-the-fly. This sounds like fun but the variable speed VFD would really need to be varying its speed all the time but the accelerations and decelerations would be very smooth.

 

[OkiePC]

Peter, can you refresh that link to the video, I missed it yesterday, and the link isn't working for me today.

Thanks,
Paul

 

[bkottaras]

You can download a sample program from the aB web site.
Look for "Smart Belt".
It was there a while ago not sure if it is still posted.
Although they use servos you can adjust your logic accordingly or at least get the basis for your set up.

 

[OkiePC]

http://www.ab.com/motion/applications/smartbelt_kinetix.html

http://www.youtube.com/watch?v=0XWboppiKi4

http://www.youtube.com/watch?v=ylAG9Ky7cxo

 

[Peter Nachtwey]

Peter, can you refresh that link to the video, I missed it yesterday, and the link isn't working for me today.

Thanks,
Paul

I don't know what you mean. It works for me. Right click on it and save it to your disk then play it.

The smart belt is simply having the down stream belt go faster or slower. I so no magic in that.

 

[Mickey]

I don't know what you mean. It works for me. Right click on it and save it to your disk then play it.

The smart belt is simply having the down stream belt go faster or slower. I so no magic in that.

It worked for me, I didn't have to save it.

 

[OkiePC]

It's probably my brow-zer (Chrome) and/or WMP codec, but the error said "Network File Error", so I thought it was a temporary link or something.

Right click, save as... appears to be working...thanks.

 

[OkiePC]

The smart belt is simply having the down stream belt go faster or slower. I so no magic in that.

Yes, they use two servo belts to gain control of the timing coming in versus going out, but the "tray conveyor" speed appears to be constant, much like the O.P. situation. That looks like a pretty simple to program way of using two servos...trying to do that with one belt would be more challenging...

Okay, finally got to see Peter's video. Very nice. Very smooth. Low wear, high repeatability.

Are the linear axes (opening and closing the gap) hydraulic on that machine?

 

[Peter Nachtwey]

This is a pure servo motor application, no hydraulics. The gates are belt driven. You can see one of the motors in the video but the frame rate isn't fast enough to catch the smooth motion. All the servos were AB and if I remember right they were 15HP. There was a PLC5 that communicated with the motion controller and monitored how many sheets of plywood were dropped. The PLC5 also moved the billet down below. After the plywood billet was complete the PLC5 controlled a conveyor that moved the billet into a press. The end product was a very long 4x12

This demo will be at a AISE trade show soon.
http://www.mtssensors.com/news/index.html
The dart position is controlled by a human in the manual mode and is therefore random.
The balloon makes moves relative to the dart position, which is always moving, plus an offset so the balloon just touches the dart without popping. It really doesn't make any difference if markymark31's fixed speed conveyor was a variable speed conveyor, markymark31 could still drop his widgets into the tray if the tray speeds were changing and the spaces between the widgets or the trays are uneven.