Which routine vibration condition monitoring program scenario do you think will give better unscheduled downtime avoidance on critical rotating machinery?

% Experienced Condition Monitoring Engineer collecting vibration data, sending data to remote location for other experienced Engineers / Technicians (with knowledge of similar machines) to analyse.

or

@ Experienced Engineers / Technicians collecting data and analysis performed by themselves or other experienced Engineers / Technicians who, take turns / regularly collect data, from the same machines.

Simply put, does a service that separates the data collection task from the analysis task produce better or worse results compared to one that does not?

Could you avoid "depends" answers please indicate % or @.

Thanks in advance.

John

This message has been edited. Last edited by: Vibe-Rater,

Naturally the best is: experienced qualified analyst collecting data and analyzing it. But generally confined to the relam of consultant.

With good techs gathering dependable data, you only miss-out on the "being there" aspect.

Sending in remotely isolates but can be good if the program is very good.

Or, my take on it anyway.

John
I like the idea of having both aspects of this type of program to be done by the same trained and qualified person.You know the see, feel,and touch that one gets by being on the spot.while both scenarios are viable.The time lost in the transition of data could be the difference between running an not.If your equipmnet is in goood order all the time remaote services can be used and have excellent results.But for the we need to monitor a machine right now and for the next two weeks before a shut is possible a man inside is the way to go.I am slightly bias as I am the man inside.So you can take from this what you will as I am sure there are people waiting to tell the other side.Hope this helps.

I don't know how anyone could put a "%" or a "@" figure since I doubt there are many analysts here that do both in house and remote/external data analysis. I have an in house program and have sent data out on occasion for analysis. But, this is usually due to a manager request for a "second opinion". The opinion which comes back with all sorts of disaster scenarios given the individual has no idea what he/she is talking about. Now, I am sure there are companies out there that claim a high level of accuracy with the data dog - remote analysis set up. The main reason being the fact that it is a business that exists. Business exists to make money for the owner of course I won't dispute that. A good business it would be where you look at data and form opinions. Now, remember as an experienced analyst (10+ years) I do most of my analysis AT THE MACHINE while collecting data. This is a fact that will be confirmed by many on this forum and is not a "depends" thing. This immediate, full system analysis is impossible with remote data analysis, sorry. So, that would be stike one for them.

I believe that a great deal of analysis is not not with time waveforms and FFT, but with eyes, ears, nose, feet, etc. While there are scenarios in which sending data out for analysis might be applicable, such as a complex gearbox analysis, the best vibration data analysis is that which compiles data from numerous sensors - including the human ones.

Just my two cents worth...

I'm with Stan. Use all the tools in the box and there's no substitute for actually being there.

I think the remote programs can work effectively for finding routine problems on most machinery, providing quality measurements are taken at the exact same locations and under the same test conditions every time. Most of the machines we look at are pretty simple to analyze and are not the things great case studies are made of.

When it comes to the odd-ball machines requiring more extensive efforts to troubleshoot, the remote approach will be ineffective; a change in vibration levels and/or pattern should be detected by a remote program, but determining the cause can be difficult. As others have said, when troubleshooting, there is far more information to look at than just the vibration graphs. All of the analyst's senses come into play. The strange noises. The vibration in your feet at certain locations. The things you can see. Even the smell of a burning belt. I know for me that some time spent "noticing things" can really pay off later. There are often indications that may seem of little importance until I'm looking at the graphs later. Suddenly the clicking noise I heard or vibration I felt in the piping suddenly becomes clear. Little signs that didn't even seem to have anything to do with the machine being tested may be important to solving the problem at hand.

The downside of a remote program with a complex problem is reaching the wrong conclusion because important pieces are unknown to the analyst. They will provide a diagnosis to fit the data they have, which can lead to extra work and unnecessary repairs in some situations.

Jon
Spintelligent Labs

For remote analysis I would vote for this board

Standing near the machine gives more clues than a remote analysist or "Expert Computer System" would ever dream of.

What I experience, a technician take the readings for offline vibration monitoring and inform an engineer of any abnormal readings for analysis.

If the organization is new to vibration monitoring and no inhouse expert yet, may be an external vibration analyst is needed for data interpretation after the technician being trained to take readings. If the machinery is critical & further troubleshooting is needed, the external vibration analyst can be called in for closer look. This is also maybe for training the inhouse engineer on vibration. Even sometimes, it may be necessary to cross check an inhouse analysis with an external expert eg before shutting down a critical machinery.

However, if you are talking about critical machinery without a hot standby, then an online vibration monitoring should exist. I think portable vibration collector should not be used for a prolonged period in case.

This message has been edited. Last edited by: Josh,

I think the time has passed that a tecnician is merely regarded to take data and bring it to the engineer sitting on his b.. behind a computer screen.

That's good if your technician can collect & inteprete data. So how many are level 1 or 2 vibration analysts?

Hi guys,

This is a difficult question to answer with numbers or percentage!! Infact, i'm doing both-taking my own data and analysing and doing analysis for datas taken by some other guy.When i take a data, if there is any abnormal condition, most probably i will make up my mind on what to write in the report in the site itself(Everytime it will not work ). But if it come from some other guy, i have to ask lot of questions on the sound they hear, how the structure behave and all.If i get good answer, i may be able to make better anaysis.

Have a nice day!!

Hi,
I have instituted a progam in which there is a group of technicians for recording routine readings. They can do 'basic' vibration analysis (maybe a level 1/2). Actual analysis of problems are done by engineers who go back to the abnomal equipment are carry out further analysis.
This way my failures without warning is low and the analysis accuracy is high.
A bit of % and @ so to spk.

Deepak

I too am a believer of who takes the data analyzes it.
I have been put into both sides of the argument with the company's I have worked for in the past.
There seems to be alot of extra testing required to get a desired result for the client with an external person doing the analysis. This does not happen every time but there is generally alot of two and throwing of information and sometimes as we all know time can be very critical.

Give the data collector some goals and a career and he or she may just want to stay with you a bit longer. Training is the KEY

Now that would help our skills shortage somewhat don't you think!


Just my opinion

Cheers

Hooch

But we have to realize: the program should go on and last 'forever' while the analyst may move on, retire, etc. Same goes for the tech; he/she may move-on to somewhere else. In the union environment 'hands/techs' may bid out or change often and you're locked-in to a program where you can't touch the data acquistion task.

Be consistant: even software isn't always consistant nor is the signal processed the same in every black box even though it should conform to a standard some want to invent their own.

I agree with both Jon Chandler and Sam. I have been assisting clients to set up and run PdM programs for 15 years now. It is very clear to me which programs will succeed, which will fail and why (you get it after the first 50 or so clients...).

It may be counter-intuitive - but the facilities that have the "on-site expert" who is great at troubleshooting machines, but does not manage, maintain, define, document or organize a coherent monitoring program - ineviatbly have little success.

The sites that have a very well organized, methodical and well documented approach - but little "expertise" are typically much more successful.

There are alot of reasons for this, to understand them one needs to ask "why do these programs not live up to expectations or fail?". There are also technical reasons why well organized programs fare better - and why "trending" is more cost effective and accurate than "troubleshooting"...

If anyone wants more info, let me know - I'd be happy to expand on this. I know there will be alot of disagreement but that is because alot of people on this board are the "experts" and obviously (and rightly) will defend their expertise and its value. If they use that expertise to develop a program that will be just as successful without them as with them then they truly are the experts and I would hire them any day! If not, well... :-)


www.DLIengineering.com