Just because they are engineers doesn't mean they have the ability to assemble the unit. I worked with one that admitted that he could not put something together but he could figure out how it should go together.
I was a little curious about the background of their experts too. The magazine stated that they were engineers working in the agriculture industry and that was about it. Dust control is a huge issue in the agriculture industry, particularly in the storage of grain. I wish the Wood article had said something about it if that were the background from which they came.
What I find encouraging is that quite a few people are questioning the results- to quote Shakespeare (who does that?) "there is something rotten in the State of Denmark." Not all that many years ago the results wouldn't have been challenged hardly at all.
Still doesn't sound like a problem of test design or engineering but some mechanical or physical issue with some piece of equipment. System was set up and the data interpreted but no one stepped back and said " these results don't seem to make sense". None of the cyclones differ all that much and if you look at all the blower tables from Cincinnati, New York, or Chicago Blower and compare the ratings for similar impeller designs with different housings you won't find all that much variation among BI or BC blades. Wynn Nano filters or Oneida HEPA rated won't show much difference on a Dylos if rated for the delivered cfm. Differences in longevity should be more significant than up front performance and that never gets tested. Dave
By the way, I find it interesting that so many people intensely question the credibility of the Wood Magazine experts but nobody is questioning the credibility of Bill Pentz, who designed the Clearvue cyclones. So far as I know, he does not have any formal training on the subject. Contrary to what others have said, cyclone design is not simple and obvious. He has published a lot of good information but he has published some misinformation too. He is not just an independent authority on the subject. Several years ago, Bill was asked on this or some other forum whether he had a financial interest in promoting the Clearvue brand and he said it was not relevant and was nobody else's business.
Because we didn't hire Bill Pentz for his expertise, so his expertise is completely irrelevant. We bought Clearvue cyclones, have experienced the performance first hand, and some of us have measured the performance, and from both experience and just simple logic, the magazine results are very difficult to believe.
You obviously have a problem with Bill, and you've said so in the past. Why not start a different thread? This has nothing to do with Bill. It has to do with the physical performance of dust collection systems. If the magazine results were reasonable, no one would question them.
Like years ago, there is still a lot of playing loose and free with the DC numbers.
I was just reading the literature for a system from one manufacturer whose website claims their "949 CFM" unit was "HEPA" which is really a filter dependent property. It was kinda weasel-worded so I couldn't tell what that actually meant, but they provided a link to "test results." The "results" were actually an abbreviated promotional spec sheet for the GE PTFE membrane used to coat a backer to make various types of HEPA filters, not specs for a complete filter assembly. The membrane had an air permeability rating of only 4 – 8 CFM per sq. ft. at .5" of static pressure drop. I doubt it is linear, but maybe, but in any case do the math to see what that does to the the claimed the 949 CFM with a 100 sq. ft. filter. At .5" of SP, the membrane alone will only allow 400 - 800 CFM. And how many systems are operated at .5" SP? I suspect even a portable unit, connected to just about any WW machine with minimal flex and an elbow connected to the supplied HEPA filter, generates more pressure drop than .5". Finally, the fan table (they don't include an actual fan curve, at least I coluldn't find one) shows a CFM of over 500 CFM through 4" duct. At a typical(?) DC air velocity of 4000 fpm, 4" duct will only pass max of 349 CFM! Even at 6000 fpm (possible?) a 4" duct will only pass a max of 524 CFM. Add the pressure drop from a machine, pipe, fittings, filters, etc. and what are you left with?
Last edited by Alan Schaffter; 01-15-2014 at 3:43 PM.
That's easy. Because truly impartial experts would not have left it at "unit B passed 20x the fines of previously tested unit A," they would have explained WHY that was occurring.
There are really only two possible explanations that I can think of (anyone reading this is welcome to offer others): (1) Inadequate filters. (2) Leak.
So let's employ the process of elimination...
If you think the filters are inadequate, just say so and provide a little documentation (there are well-defined standards for filter media) and you're done.
Excluding the filters, it has to be a leak, right? Just fix the leak and retest. Smear every seam with a gobs of silicone if need be. Get another tube of caulk, if necessary, make it ugly. And you can mention this in the article ("man, that thing had a leak and it was a bear to find").
If you have smeared all the seams, and the problem continues, it has to be a hole somewhere in the filters, right? They could have swapped the filters between unit A and B. They could contact the manufacturer and request a 2nd set of filters.
Those are the steps I believe a true expert would have taken.
Leaving the reader with the impression that unit B passes 20x the fines, and that this is typical of all unit B's, not so much.
All just my personal opinions.
Good observations, Phil.
One can never have too many planes and chisels... or so I'm learning!!
Phil. Read the article.
I picked up a copy of wood after work today and read the article. I'm not a Wood subscriber so I can't go back and read the article this is a follow up to.
After the first two sentences I already knew this would have been considered invalid results at the school I went to. I say this, because the article reads as if they picked up a Clear Vue, tested it and then re-plotted the results on top of the results from 6+ months ago. also based on the article in the photo, it looks like the test environment is just a random shop. Some random shop doesn't provide a very stable test environment.
My thoughts based on how I read the article:
1. You cannot compare results from two tests separated by this much time, in an environment as unstable as a work shop.
1.1 The shop could have more dust lying around on shelves and benches from projects built since the original tests. Thus the tester can stir up said dust, and skew results .
1.2 The dust generation source (drum sander?) might be in a lessor state of repair than it was during the original tests, and thus could put more dust into the air and skew results.
1.3 the ducting used could have the same issue as the drum sander.
1.4 the calibration of the testing equipment could be off if it wasn't re-calibrated.
1.5 who conducted the tests this time? the article doesn't say if it was the engineers that people on this thread did the original tests, or some wood magazine employees, just trying to duplicate what the engineers did originally.
1.6 what was the sample size?
1.6.1 was a big enough sample taken so outliers could be excluded and a statistically sound average generated per machine tested?
1.6.2 was the data validated with say a chi square test?
It's been over a decade (undergrad) since I've done research, but even the few things i listed above off the top of my head, would have gotten a lab report tossed in the trash.
-Dan