Troubleshooting a Steel Rolling Mill

Ask where the quality assurance document/standard manufacturing practice document is for the process, and start there.

If the facility doesn't know how to make the stuff correctly, then how is an interviewee supposed to know?

Oooh! Just thought: use the PFMEA and Control Plan!

I've spent enough time writing these darn' things over the years...let's put 'em to good use I love Quality Systems....

<waves a feather on a string and opens a tin of tuna to distract Del...>

Yep. Why have fifteen duds been made instead of just one?

Are all fifteen duds the same?

Solve the issue on the one, and the other fourteen will go away as well.

Agreed. Also, the distribution of failed coils with time.

Got it! Thanks

Good point! (and GA)

I'll ask about the Archery. It's about time we had a shoot-off

What about a metallurgy sample on the material before it enters the rolling mill, to make sure it meets spec?

Also, it wouldn't hurt to re-check the analysis to make sure the finished goods are out of spec. There's always the chance of an error by the technician doing the analysis or that the tools he/she uses are out of calibration.

GA. There's another slide - I reckon 10 should get me through the 15 mins.

I've been told to allow 1.5 hours for total interview "experience"!

I'd leave this statement, "Oftentimes a root cause will have a root cause" out of anything you present to the interviewer.

I didn't say that's how she should present it, just something to think about.

Thanks lyn...checking it out now

I'm not sure how detailed your presentation needs to be. For the first part, describing how to find root causes, you might want to consider explaining your strategy. How did you come up with that list of potential root causes? For example, are you using lean manufacturing techniques to be sure nothing is overlooked? Oftentimes what at first glance seems to be a root cause will have a root cause, so a technique like the 5 whys analysis will help you find the true root.

For this type of interview your process can be just as or more important than your final answer. They can train you on the job but not how to use your brain!

(Sorry, I replied earlier under the wrong user name!)

I wondered how lyn replied in #5 to post #7....
Good point on the 5 whys - I use the technique but it's "so obvious" I don't need to explain it...LOL
Hmmm....slide on 5 whys & 5C fault analysis....slide on possible causes....slide on WHO to ask for more details.....slide on tests to be carried out.....slide or two on effects on company (demonstrate commercial awareness)...Feeling better about this...

(Sniff...) I smell conspiracy!

Is there a pattern to the fifteen; were they all evening shift, for example?

<slaps forehead> Oh yes!

Funny, at the last place, that was the first (or maybe second) question I asked...didn't occur to me in this pretend scenario!

Sometimes intelligent people overlook the bleedin' obvious. Don't be frightened to ask the question. The only stupid question is the one that is never asked.

Were these LynDoor™ coils? Nuff said...

Could be.

We use the Microsoft manufacturing philosophy. Get it close enough to ship and let the end user sort it out.

Here at LynDoor™ we subscribe the the 5 W's quality plan.

1. Why does it need to be that good?

2. What difference does it make?

3. When did that requirement get added?

4. Where did that spec go?

5. Who cares?

5Ws and and H...

How do they get away with it??

Careful, there. KrisDel™ might action for plagiarism.

Yeah, very careful - I think Hillary Clinton has that copyrighted.

Just thought of another: Check maintenance records. And training records, possibly.

"Check maintenance records." If there are any.

One day it was working correctly, now it isn't. What changed? Were the roughing or finishing rolls changed? Pickling solution?

As far as the surface finish, having a series of diameter changes with the wrong gap or speed settings can cause defects as well. Including micro-cracks toward the interior.

"control system issue perhaps? " ARRGH! Sorry, you just hit a sore point. If you mean the instrumentation, or I/O then OK. Most people (read OPS) mean program when they say control system. A DCS/PLC very very rarely goes "sort of bad". If it could add 2+2 and get 4 yesterday, it can do so today, or it is an obvious molten heap sobbing in the corner. If someone has changed the program (again what human changed what), or mis-tuned a PID loop, that is a different story.

2 cents

Thanks for the nudge - I did indeed mean instrumentation, I/O or damage to the PLC.

I'm a mechie, so the boxes "do magic" and are always right (from their input to their output). Honest

Duh! 8D Analysis (for some reason, the last company at which I worked called it 5D - I know they left out D8; I'm trying to work out which other three* they omitted...)

*There are actually nine stages in the 8D process - D0 to D8....

I have a suggestion for D9. Teach your team to count. Or, is that D10?

Perhaps it was designed by a computer scientist...

Hang on, no. Then it would be to 0101D Process....or the 09 0D Process

I'm confused as to how/why you Plan (D0) before putting together your Team (D1)

More likely an MBA. Those who can, do. Those who can't, become business majors.

No, this is D10.

LOL!

A single point to start and concentrate on: What component or equipment is responsible for the "surface finish"? Approach faulting in discrete steps.

One of the questions I have!

It seems logical that the surface finish of the rollers will have an effect, as will the pressure load and the thickness reduction. Anyone know more?

Sorry if I missed it, but are we talking of Hot Rolling of steel or Cold Rolling ? It definitely makes a difference.

It's not specified in the interview question...so I "decided" it's hot rolling so I'll mention temperature control and keeping it above the recrystallisation temperature.

One last thought: You mentioned "temp survey needed". Along those same lines (whether this is hot or cold rolled), what type of historical data gathering do they have for doing analysis on the process. If they have a SCADA system, they will most likely have some sort of historical logging package (PI, INSQL...).

Process knowledge, raw data, and intelligence make a might powerful troubleshooting combination.

Great minds!Slide 7: Temperature survey (i.e. what are we getting) and review of logged data (when did it (if it did) change)
I like to think I have the intelligence (or can tap into others'!! Slide 3: Assemble Team), I think I've covered the raw data....wish I had more process knowledge! I'm working on the principle Savvy mentioned - they are more interested in how I think that what I know. Of course Slide 3 is also about using more knowledge than I have.
Thanks for taking the time to suggest. I hope maybe the thread will help others preparing a similar presentation.

A huge thank you to everyone who commented here. The interview went pretty well today (0900-1030). Turns out I'd met the interviewer when he was a baby engineer...and I didn't remember, although he did! Generously, he only told me that on the way back to reception

About three weeks until I hear back (I'm the first they're interviewing), so fingers crossed.

Suggestion: follow-up letter/e-mail/whatever, thanking them for the interview opportunity and selling something else: "I don't remember mentioning that I......[insert something to savour].". Something along those lines.

Bugger - I thought of something on my way home....can't remember what it was now....will sleep on it...

Thanks