Magneting data recording on steel

[Prince]

The pipes go to differnt branches due to the result of the tests, I think the idea of counting can not be trusted in this application. I am still very interested to know about any possible circular marking....

 

[jstolaruk]

I routinely use Carco for ink marking systems. They have different typee and colors of ink and will tell you what works best with your material after sending them a sample. I'm confident they could work with you on a marking system.

I worked for them integrating some color sensing systems back in the late 80s early 90s sold to Ford when color sensors were just making an appearance albeit very expensive. Now there are quite a few mfgs os sensors and prices are relatively cheap.

 

[elevmike]

Prince,

I just dreamed up the idea as the result of this thread. It's a very doable concept. The marking and color vision arent the big challange here.

The big challange is time!! You need to mark up to 2000 20" diamater pipes in one day, then get them through the entire inspection process within a reasonable time. So the motion control is the challange. You need to design hardware that's going to complete the marking in time to finish inspecting the 2000th pipe before the end of the day.

What is the mass of the largest pipe?
What is the max dwel time allowed for the pipe to be marked an compleated inspection? Etc... If your day is 8 hours and the inspection process requries 2 hours for each pipe, then you only have 7 secods or so to mark each pipe. That might require some pretty robust motion equipment for moving and rotating large pipe.

 

[elevmike]

you definately want to look at the link jstolaruk posted. Spicifically here: http://www.carcousa.com/non-contact-marking.html

Reading the colors: this is just a sample of how you might read the stripes: http://www.keyence.com/products/sensors/rgb/rgb.php

 

[elevmike]

Note: I just 20 seconds at each inspection station. So if your inpection stations are linier and lets say every pipe will have to go through at least the 1st inpection station, then you have 20 seconds to mark the pipe. That's about 11+ hours for marking 2000 pipes?

 

[Lancie1]

Merely an Attempt to Think Outside the Box . . .

Problems:
Outside marks could be rubbed off.
Outside marks may not be in right position to read, without positioner rotation device at each inspection station.
Magnetic markers may not be easy to read on outside of pipe due to magnetic interference from pipe itself.

Thoughts:
What about marking the ENDS some way? Make up some "inserts" (plugs, stoppers, forms) that will stick into the end of a pipe, with a barcode on each one. This insert fits flush, no projections, but it should have two or more holes, to be used for removal at the end of the line. At the first inspection station, the operator sticks one of these into the front end of the pipe and also enters the tag number into the system database. As the pipe enters the next station, a multi-directional bar code scanner mounted just above the pipe, scanning down and into the pipe inner diameter, reads the label and knows which pipe it is working on.

Don't have an operator at the first station? Well, there are robots that can do this job. You probably would need a second one at the last station (or at the shipping dock) to remove the tag inserts.

 

[elevmike]

Lancie,

That sound like a good idea, but he could be dealing with up to 2000 pipes in one day. Additionally pipe sizes range from 8" to 20" dia. That would require, what? 20,000 stoppers maybe? + the cost of two robots, and moving the stopper back for the next days run etc.

I think the right ink/paint will adhear well enough to be seen or detected by a good color vision system. The vision system might be looking at 1/3 the dia of the pipe. Somewhere in that area the paint should be visable. There are also lots of inks to choose from for fast dry & applying to oily surfaces etc..

Here's the link on pigmented inks: http://www.carcousa.com/marking-inks.html

If Prince goes with ink marking he will likely require the XX-whatever type ink (10 second dry time)

 

[Doug_P]

That might require some pretty robust motion equipment for moving and rotating large pipe.

If the outside is to be marked, it might be easier to rotate the marking equipment around the pipe (maybe a robotic application?) than rotating the pipe itself.

 

[elevmike]

If the outside is to be marked, it might be easier to rotate the marking equipment around the pipe (maybe a robotic application?) than rotating the pipe itself.

That's a good thought. Maybe it wouldnt even requre a robot. Just spin the paint head 400 deg. on a circular track then return in the oposite direction.

My orig thought was to apply the paint from underneath the pipe. That way no matter the pipe dia. the stationary spray heads would always be the same distance from the target, & fewer moving parts, hoses etc...

 

[Doug_P]

Maybe it wouldnt even requre a robot. Just spin the paint head 400 deg. on a circular track then return in the oposite direction.

I was thinking of the 8" to 20" diameter requirement. A fixed circular track wouldn't be the same distance from the work for all possible diameters. Wouldn't have to be a robot, just something adaptable to varying pipe sizes.

 

[Prince]

I think everybody is going to something too expensive. this is not a car making company. I think more important than the problem of How I can do it. It is the problem of How expensive it would be...
But anyhow thanks for paying attention to my problem. I still think that unless I can find a marker which can mark on everyside of the pipe on line, I can not think of marking...

 

[elevmike]

Prince,

You didnt answer my questions.

What's the max pipe length & weight?

Additionally:
How many inspection stations?
Are the inspection stations automated?
What do they inspect for?

Is there a branch-off for each inspection station?
Are there inspections after ther branches? Or are the branches rejects?

You need to define your process in detail.

After some thought I came to the conclusion that it's still possable to track the pipes without marking them, and even if there are branches. (Only so long as one pipe cant get ahead of another in any given process/branch).

Expense is a relative term. Figure it this way:

If your tracking 2000 pipes a day x 260 days, that's 520,000 pipes a year.

What's the expected lifetime of the process? Just to guess maybe 5 years?

That's 2.5 million pipes or more.

So what's it worth to you to mark each pipe?

If you think spending $.50 on each pipe marking sounds like a deal then your budget would be in the neighborhodd of $750,000.00!

If your marking 16-20 ft lengths, I cant see spending more the $50,000.00 on the marking machine. The readers might cost $5,000 each by the time you figure in the material & labor to get them installed and running. The remainder of the budget would be for consumables & maintenance.

Before you consider any tracking system you need to ask the higher powers what it's worth to them on a per pipe basis so you can define your budget.

 

[Prince]

Dear Mike

very nice estimation that makes me think again about the rejected idea of tracking by stack again. may be I will start another thread and talk about the idea cause the system is rather complicated from station to station. but if there was a way for my main idea (recording magnetically on steel) it was not that expensive

about the questions:

What's the max pipe length & weight?
12m and 1000kg (the length won't change)

How many inspection stations?
Are the inspection stations automated?
What do they inspect for?

There are about 20 Places were I need to gather information
most of the stations are automatic
pressure, ultarsonic, eddy current, full body testers for exploring defects in pipe body or weld

Is there a branch-off for each inspection station?
Are there inspections after ther branches? Or are the branches rejects?

No there is a main line for healthy pipes but due to the standard they are producing the pathes change. some lead to other ways if fail some tests then repairs done to fix the pipe then comes back to the main path. some leave for sampling and come back to the path.

 

[elevmike]

12m = 40 feet. 1000kg = 2200 lb. Huge pipes!!

So the branches are ejects due to defects? When the pipe is repaird I would presume that it would start at the begining of the line? I'm wondering why trak them at all?

Tracking by FIFO stack:
You did'nt mention brand of PLC but the general idea is not really as complicated as it seems. If you have a max 20 pipes in the process then maybe you only need 20 memory locations in the stack. A counter or indexer would count the pipes comming into the system. Each station would be represented by a single memory location that would contain the value assigned to the pipe in that station. If a pipe gets ejected from that station then the value representing that pipe would be moved to another stack for tracking the pipes ejected from that particular station. If the pipe is moved to the next station then you move the value down the stack by one memory location and so on till the pipe reaches the end of the process.

To decide exactly how you manipulate the memory stack you need to graphicly draw the process flow from station to station or ejects etc..Then imagine just one pipe going through the entire process.

 

[jstolaruk]

but if there was a way for my main idea (recording magnetically on steel) it was not that expensive

If possible, recording may be cheaper than ink, but how were you expecting to read it? I would expect the magnetic reader to be a lot more expensive than the color sensors.

I would do a full cost analysis before dismissing any ideas proposed here so quickly. Then again, if management hasn't worked up a budgetary number (based on a cost-benefit-pay back study), you're probably swimming up-stream anyways.