Automation of a Coating Process

[arpus4KM]

Not knowing the actual process/ hardware that exists etc. I think people may be over thinking this, most dipping processes I have come across the tanks are large enough to take the components with spare room all round, if this is the case then why go to to the extremes of hi res encoders or other methods that cost a lot more, difficult for many maintenance staff to faultfind on. sometimes simple is best.
On a side note: the coating process how is the PH controlled I assume this is manual at the moment.

I don't know about overthinking it, but being mindful of the type of process.

I've had to deal with the aftermath of some crane accidents, that those were the result of an operator not paying attention, but then noticing. Now remove the operator from it and let the process run on it's own without enough feedback.

 

[parky]

I was not saying make it unsafe, indeed hard wired limit switches is probably more reliable than a radar sensor or other electronic means, as for the accuracy (if you read the post properly) I mentioned (as I have experience) the tanks are usually wide enough to give plenty of clearance so a mm or 2 is neither here or there, as I mentioned on my first post, alarm conditions to stop the process are a must. I know quite a bit about coating processes, the chemicals used & makeup/ disposal. I do however find the use of a chain driven hoist is not the ideal choice but without knowing the clearance of the system depth of imersion etc. it is difficult to give best practices on what equipment to use, in all the ones I have done including zone 0 hazardous areas the dip length is usually short but it will depend on the size of parts & the mounting depth.
I'm surprised that there is no mention of the tank conditioning/filling process, the OP stated that the operator will be employed elsewhere, it would make sense to automate this as well,
the treatment is chromium free passivation on aluminum, zinc and some other metals.
The tank prep should be half fill with deonized water, adjust the PH, add the chromium coating solution, top up with water check & adjust the PH. it should be below I believe 4.0, there are also other things like constant agitation, heating etc.

 

[eaglewood]

Not knowing the actual process/ hardware that exists etc. I think people may be over thinking this, most dipping processes I have come across the tanks are large enough to take the components with spare room all round, if this is the case then why go to to the extremes of hi res encoders or other methods that cost a lot more, difficult for many maintenance staff to faultfind on. sometimes simple is best.
On a side note: the coating process how is the PH controlled I assume this is manual at the moment.

I agree Parky that simple is sometimes best. And you're right about the clearance. I've got 1.5" clearance all around the square baskets that will be dipped into the containers. Millimeters of accuracy not required.

I was ready to go with the arrangement you mentioned in your earlier post until I read Gary's post. His mention of it lasting 20 years grabbed my attention. The laser distance sensor struck me as a good idea to keep from installing about 20 limits (for 8 dipping containers, home, and end stops). And my runway being only about 19', I can get a laser sensor for under $500. SEEMs like the way to go, but I'm holding back from pulling the trigger, while I learn from you veterans. What a valuable site this is. I'm it here in this little factory, when it comes to automation and controls, and now it's like I've got mentors. Can't tell you how much I appreciate it.

The $500 laser I’m considering (https://www.automation24.com/laser-distance-sensor-ifm-efector-o1d100-o1dlf3kg) outputs 4-20mA, and it doesn’t seem too difficult to scale and calibrate. The baskets in the process are all equidistant. It just seems one of these would be simpler than the 20 limits, though I will still have the end stops and maybe home.

It sounds like I have a choice between a star wheel and an encoder to control the hoist dip length. And you’re right about the short dip length, parky, it’s about five or six feet. I can also imagine controlling the speed of the hoist and then clocking how long it takes to arrive, and put that timing into the plc, but that might be asking for trouble. I’m back to thinking an encoder will be best for the hoist. I see a lot of incremental encoders to choose from for under $300.

When I originally balked at encoders, I was reacting to using a rack-and-pinion linear encoder for the trolley location. That would work great, I’m sure, but I imagine more headaches at installation compared to the limits/laser. Limits or laser distance sensors seem preferable for the trolley, and an encoder or star wheel for the hoist.

To answer your pH question, the pH adjustments will be carried out manually for now. That presents an automation project for the future.

 

[parky]

Yes I think you have the right approach, a laser distance seems ok, however, fumes from tanks if there are any could possibly affect it, but that would depend on many factors, the one coating system I re-hashed I did not see any real problem with visible fumes so probably ok & lasers should not be affected that much. The drop of 5-6 feet is greater than I expected, but not knowing your system I guess the parts must be quite sizable.

 

[eaglewood]

We do have nitric acid fumes but they will be ventilated. The rest of the tanks won't present a problem.

The 5-6' is if I reel it all the way up, which I guess I shouldn't do. So if I just lift high enough to clear the baskets from the top of the tanks, the drop will be 2-3'.

 

[parky]

Just as a mater of interest, attached is a re-hashed printout of a system I did some time ago, this was just a modified & stripped out version of it where I replaced the counters with limits, this is only a simple one without alarms for training an aprentice I had at the time, the program block "Sim" was used to test it on a simulator, My apprentice took a copy & I asked him to produce alarms should it not reach position & incorporate them into the program to give him some experience, all I have done is changed a few symbols to reflect closer what you may have but there are no bells & whistles so only an idea.
By the way, this uses an integer word as the step sequence, this is a good way of doing it rather than memory bits, it has the advantage of being able to add extra stages (like an actual other operation or perhaps time delays between certain steps that is why it generally runs in steps of 10). the other bonus is that most HMI's have the ability to display text based on the sequence number, all it takes is to add the list i.e. Seq_Word 0, "IDLE, 5, "Waiting for product" 10 "Bogey Down" etc. so it automatically displays the current operation on the HMI.

 

[GaryS]

I just thought I would try to clarify as few points
With a conical brake or load brake the motor must be running at some speed for them to work that minimum speed may make it difficult to do the positioning you want and are not necessary.
I would recommend not using either

As I stated before I recommend using a flux vector drive on both the hoist and trolley as you know they require an encoder attached to the motor
With this configuration you can control the speed all the way down to zero speed this will give you better position accuracy much better then you need. I see that as a plus in all cases

As for the laser there is no doubt that it the best way to go here if properly installed, I have systems running over 20 year without maintence issue.
I would avoid analog feedback from the laser as we know all analog systems can change over time and temperature and you have to scale the input most analog cards are limited to 16 bits at best some are only 12 bits. That’s not really good resolution to me, then you have the value you read in may change with temperature change resulting in an error in the actual position.
I have found the best way to read the position from the laser is the ASCII the system output a continuous ASCII string you just need to decode it the data received will be the exact distance it does not care about the ambient temperature it will be consistent as long as the laser is power and working.
The data that is received is easily converted to a real number using a function in the plc no scaling necessary.
Ethernet and modbus communication must be polled to be used but the poling time may change with each pole the more devices on the network the more delay. And that delay time may be inconsistent.
And you still will need to convert the data received from ASCII to real to be used in the plc so the read time would be longer with this method.
In all positioning system update time in important
As for never working with ASCII. You never worked with a plc until your first and you learned from that
Now is your chance to learn how to use and ASCII module the next time will be easier. We should all strive to learn something new every day.

As for using a reel, reels are high maintence items, as time goes on the brushes ware and fail causing down time yes they do have reels that don’t have brushes but those must be mounted on the hoist. Depending on the requirements you may have to use a power reel, again more maintence.
A good festoon system is the best way to go on this

I hope this helps

 

[eaglewood]

As for never working with ASCII. You never worked with a plc until your first and you learned from that
Now is your chance to learn how to use and ASCII module the next time will be easier. We should all strive to learn something new every day.

Gary, I'll be taking your whole post to heart, but the above really jumped out at me. Absolutely, I'll be learning like crazy during this project, including ASCII.

And thanks, parky as well for the FBD program. That's another thing I've only dealt a little with. I've been programming 99% ladder logic and I like it, so why fix what ain't broke, sort of thing. But I've heard FBD is powerful. It's probably time to use that as well.

Thanks, guys!