How can one measure compressor performance?

[IT_Architect]

Compressors are rated under a standard set of conditions that are not convenient for us to duplicate. Calculating the raw CFM on any particular day is not difficult. The goal is to find or establish a method to convert measured output to what it would be under standard conditions to:
1. Verify manufacturers' claims
2. Monitor compressor condition

The method does not have to be limited to a spreadsheet formula, and can leverage online tools, such as the engineering tool box and computer programs, because I realize that there will likely be elements of the equation that are not linear.

The important thing is that how and why it works can be explained in terms most people can understand.

Thanks!

 

[Techie1961]

Not really sure what you are looking for. To verify a manufacturer's claims would require a lab and set of standards to which you want to test to. There are so many variables and you would need to know what the industry standard approach is.

As an example, is the rated CFM simply a calculation of the displacement or is it to real world conditions and account for restrictions in inlet flow, outlet flow, back pressure from tank, RPM, etc.

 

[Bondo]

Calculating the raw CFM on any particular day is not difficult. The goal is to find or establish a method to convert measured output to what it would be under standard conditions to:

Ayuh,.... Shut it off, drain the tank, turn it on,...

How long does it take to fill how biga tank to what pressure,..??

 

[Kaizen]

um what? if you want to verify manufacturers cfm just install a gauge.
what the heck would you need on line tools and spreadsheets for? solution looking for a problem

 

[IT_Architect]

I re-read my question. What I'm asking for and the purpose are spelled out clearly. It is a technical question. If the question or the purpose do not make sense to you, or if it appears to be a very simple answer, then this question may be a little beyond what you are able to help with.

Thank you

 

[mike93lx]

I re-read my question. What I'm asking for and the purpose are spelled out clearly. It is a technical question. If the question or the purpose do not make sense to you, or if it appears to be a very simple answer, then this question may be a little beyond what you are able to help with.

Thank you

Your post does not contain a question.

 

[rockwithjason]

I re-read my question. What I'm asking for and the purpose are spelled out clearly. It is a technical question. If the question or the purpose do not make sense to you, or if it appears to be a very simple answer, then this question may be a little beyond what you are able to help with.

Thank you

with an attitude like that they reply you will most likely get is "screw you".

 

[IT_Architect]

Your post does not contain a question.

The title of the post contains the question. If you think I should add it to the content, I could do that.

 

[3rdgen]

I'm no expert but to monitor output could you not use a combination of flow rate, air pressure, current draw, compressor and motor temperature. Also time to reach a certain psi in the tank with no output. As for verifying manufacture claims you would need to know ambient temperature, altitude, humidity, etc to duplicate their testing procedures.

 

[Kaizen]

I re-read my question. What I'm asking for and the purpose are spelled out clearly. It is a technical question. If the question or the purpose do not make sense to you, or if it appears to be a very simple answer, then this question may be a little beyond what you are able to help with.

Thank you

First off you're an idiot. Looking at your other posts here shows you apparently tried to write this one in Yoda talk......idiot you are yes? Furthermore for a guy that claims to own an IT company.......holy ****. its pure data. to be exact probably upwards of 20 points of data with at least 4 formulas that don't mesh together to get one magic number. bottom line is maybe a mechanical machine and how it works is beyond you because you are trying to digitize thousands of different parts in thousands of different configurations for what end? For you to make an app? To buy a compressor? Just cause you have nothing to do? Next time format your question as you have in the past and drop the attitude maybe you'll get some help

 

[pancho400cid]

As the OP said... it is not at all simple.

You can use a rotameter or other flow meter. Depending on the gauge, you will have to correct for density (and by extension get an SCFM value from an ACFM reading) which means you also have to know the temp and relative humidity (probably can neglect RH for most needs). You correct the value you read using the correlation equations provided with the gauge. Compressor flow is usually published in standard cubic feet per minute (SCFM).

http://www.omega.com/pptst/FL1600.html

 

[finn]

Just take a refresher course in college level thermodynamics.

It's only math, but this isn't the site (or audience) to get too deep into answering the question.

If you want to monitor the condition of your compressor, the stopwatch and time to reach spessure seems best.

a data recorder would be great, but plotting pressure vs time (30 second intervals) should be accurate enough.

Don't forget to record ambient temp and relative humidity, barometer, and tank temperature too.


A calibrated bellmouth and a pitot tube along with appropriate pressure transducers to measure aitflow into the compressor would also be useful to improve accuracy.
Have at it.

 

[IT_Architect]

...you would need to know what the industry standard approach is.

True. I have found some information on standard test but that is also part of what I'm asking. The only ones I know are for aviation.

...would require a lab and set of standards to which you want to test to.

In aviation, performance certifications are done under non-standard conditions. Standard-day zero-MSL performance is a calculation. In non-military applications, engine manufacturers monitor their engines during operation under highly valuable conditions and can detect condition changes that affect performance down to the baseline. Coming up with procedures to refactor measurements of positive displacement air compressors to "standard day" should be quite a bit simpler, but not simple.

I have done some work on this area already. The purpose for starting the thread is to attract like-minded individuals that either have done this already, want to contribute, or have the technical skills to critique possible solutions. This would have a lot of uses. It could verify performance claims for both new or a used compressor before you buy it. I could quantify a compressor's condition so you could know what you could expect for improvement if you rebuilt or replaced it. It could help spec. the size of compressor required whether the location is 100 MSL or 6000 MSL.

 

[CNGsaves]

Let me guess . . . . this is for pancake compressor ?? . .

Where are PICS of whatever compressor you're so worried about "performance" or CFM ??

 

[IT_Architect]

As the OP said... it is not at all simple...Depending on the gauge, you will have to correct for density (and by extension get an SCFM value from an ACFM reading) which means you also have to know the temp and relative humidity (probably can neglect RH for most needs).

That is exactly the type of common sense approach that I'm looking for in a method. There are certain considerations where the impact is so small that it it doesn't merit adding the complexity.

 

[BillK]

IT,
I would think that all you would need is an accurate inline flow meter and a pressure gauge and method to regulate the final output flow . If the manufacturer says it will supply 10cfm @ 90psi you fill the tank, open the output valve till you have 10cfm flow and then see if it will keep up at 90 lbs. . If it keeps up its good to go. If it doesn't you could vary the output valve till you see what it actually will do.

That wont eliminate all variables as far as temperatures, pipe size etc, but would be pretty darn close.

Personally I am the trusting type, if I was in the market for a compressor I would trust the manufacturers specifications.

 

[Techie1961]

IT Architect, I think, and may be wrong but the standard aviation tests that you are referring to are different than what you initially asked for. In order to compare a manufacturers specification to actual, you must use their procedure. The aviation tests would seem to qualify a component or system for use, not to establish validity of the numbers published or deviation from it. Quality control versus functional testing.

If you want to know whether a manufacturer is inflating their results, you have to know what the test criteria is. Testing a cold pump versus a hot pump along with so many other variables will only provide useless data. You must baseline and standardize.

 

[rayra]

ETA withdrawn. OP's quest seems weird and inconsequential.

 

[IT_Architect]

IT,
I would think that all you would need is an accurate inline flow meter and a pressure gauge and method to regulate the final output flow .

I would like to do just use the gauge, and do it mathematically.

IT Architect, I think, and may be wrong but the standard aviation tests that you are referring to are different than what you initially asked for. In order to compare a manufacturers specification to actual, you must use their procedure.

Compressor manufactures don't get to decide what a Standard Cubic Foot Per Minute is. SCFM conditions appear to be using aviation standard day conditions, 14.7 psi/29.92 in. hg., 15C, and a relative humidity of zero.

The aviation tests would seem to qualify a component or system for use, not to establish validity of the numbers published or deviation from it. Quality control versus functional testing.

I see them as the same. If an engine/compressor cannot do the numbers you are buying, you don't buy it. If it an engine/compressor can no longer do the number it needs to, you need to fix it.

This guys seems to know alot about air compressors and how they work

Thanks for the links.

 

[BillK]

I would like to do just use the gauge, and do it mathematically.

The only way I can think of to do that is if you had a fixed orifice for the output that had a flow vs pressure chart available for it. Not sure if such an animal is available. You simply must have a way to determine the flow in SCFM.

 

[Techie1961]

Compressor manufactures don't get to decide what a Standard Cubic Foot Per Minute is. SCFM conditions appear to be using aviation standard day conditions, 14.7 psi/29.92 in. hg., 15C, and a relative humidity of zero.

No but they can select altitude, type of air filter installed, ambient as well as running temperature of the pump, tolerance for RPM, CFM at empty tank versus full tank versus tank at maximum working pressure, etc. They might even just do a simple math calculation of displacement and RPM.

 

[theoldwizard1]

Ayuh,.... Shut it off, drain the tank, turn it on,...

How long does it take to fill how big a tank to what pressure,..??

Not super accurate, but not bad either !

 

[IT_Architect]

This is what I have so far.

My mind is not sharp enough now to go back through and make sure the logic is solid. The only thing I can say is I have it to where the reaction of the numbers to the input changes make sense now. (From the little I've tried)

The SCFM is calculated from a measurement under actual conditions, and adjusted to how the compressor would perform under standard-day conditions, which is what our compressor performance specs. are based on. The altitude I got off an altimeter but you could also get it from a map or GPS. The barometric pressure is available from any weather site for your area.

I need to get some real work done for the company tomorrow to keep customers happy. When I feel sharp, I'll go back through it. In the mean time, feel free to work through to see if I have errors in the logic.

 

[Dick in Wisconsin]

if you want to verify manufacturers cfm just install a gauge.

What kind of a gauge should be used to accomplish this?

 

[IT_Architect]

What kind of a gauge should be used to accomplish this?

Just an accurate tank pressure gauge. I know it's a little tough to know what accurate is. I just bought a new, theoretically certified, gauge because I doubted the one on my compressor and it turned out they are within 1 1/2 lbs of each other at 150, so I would guess things are not too far off. Knowing your cut-off pressure compared to gauge might also give you a hint as to your gauge's accuracy.

The particular unit that I took the measurements on is used, and specs at 16.9 CFM at 40 PSI and 16.0 CFM at 90 PSI. An average 16.12 from 0 to 152.5 might seem logical, but one cannot assume the formula is good from that.

I threw it out there for a peer review by like-minded individuals to find fault with the logic and and offer improvements. I know there are people out there more skilled than I at this. For myself, I have a business to manage, and need some space so I am able to return to critique the logic with a fresh look, and document the logic so in the future I won't have to try to figure out what I was thinking at the time.

 

[sberry]

I could quantify a compressor's condition so you could know what you could expect for improvement if you rebuilt or replaced it. It could help spec. the size of compressor required whether the location is 100 MSL or 6000 MSL.

It is an interesting exercise, if one worked for a compressor builder but what would be the point of all this? A 6 pack fueled mechanic could get so close with a guess that you couldn't separate this with a slide rule.
None of this would help with sizing or service.

 

[IT_Architect]

None of this would help with sizing or service.

That's where we disagree. Also, some people get a satisfaction from learning how things work. There are many busy sites on the internet devoted to just such things.

 

[CNGsaves]

Pedagodgical goose chase.

Who tests a compressor at 97% humidity ??

 

[Falcon67]

I think there is a bit of fallacy in relation to the aero references. We use changes in RH, water grains, etc to figure DA and effective HP at the drag strip in an attempt to pick an ET for the next round. Similar I imagine to engine power calculations for aircraft. However, we're using the oxygen content of the air to burn fuel, where a air compressor doesn't do anything but move air. The only thing I see that applies would be an intake restriction because displacement in the cylinder is displacement, less any leakdown past the rings. It pulls in a certain volume and expels a certain volume, how that volume relates to SCFM would be about the same as figuring density altitude.

Agreed this is mental ************. As long as a compressor runs my tools, I give not a **** what's on the name plate or if it meets specs precisely. All of them say pretty much the same thing anyway, +/- .5 SCFM. The only thing that has been regulated is the motor HP, to get away from "6.5 HP!" from a motor on a 120V cord LOL.

 

[sberry]

That's where we disagree. Also, some people get a satisfaction from learning how things work. There are many busy sites on the internet devoted to just such things.

I will agree with this, it is a great excercize, great for understanding, and all worthwhile except for general sizing.
There is so many factors involved that figuring it down to a fine science does not help the choice a lot.
We are using real factors of 3 to 1,,, 2x, 100% more needed etc. No one is sizing a comp because of a + or - some infintismal gnats azz of a couple % points.
As I said, the concept is great but the reason isn't for end user sziing.

 

[DekeT]

Pedagodgical goose chase.

Who tests a compressor at 97% humidity ??

That was brilliant!

 

[sberry]

Some math, some hp figures are useful as well as some experience to say that 2 hp is underpowered, 3 is marginal and requires management and 4 is adequate.
If one is trying to determine which 120V comp is better to run a shop, the difference between them all being a cfm at best is futile.
A real comp makes 12 or better with 60 to 80 gallon of storage and 2 stage for real shop work. A human works about 60 % most of the time, it can go to 100 for short periods, we need to figure the capacity at about 1.5 for 2 men till we get to industrial averaging of 3 or 4 to 1 depending on conditions. Some demand factor may be as low as .23 but if 100 people in a Boeing plant use air tools we need to provide service for 23 of them at any time.
If a single man is going to work hard with a sander he needs another 25% over the minimum of 12 I quoted and at 25 cfm can well work 3 or 4 men if we are not sandblasting.
I admire a fine mind for fine calculations but air is super wasteful for power transmission. The best waty is to get more than enough with a healthy margin.

 

[sberry]

There are some tester guys making vids, they do a super job, they should be teachers and I watch on occasion to learn but it doesn't do much for me to picking an impact wrench. All that info and I am attracted to the demo where a gearhead says he is using it in a class 8 truck shop and it works for every bolt he ran it on.
Showing me how he calculated it all is rather irrelevant.

 

[Kaizen]

What kind of a gauge should be used to accomplish this?

http://www.omega.com/googlebase/product.html?pn=FL-505&gclid=COe45Kbnl8kCFYI8gQodiEIG0A

I'm sure there are more accurate ones out there for more money. This one has error rate of 2%
http://www.grainger.com/product/HEDLAND-Flowmeter-4NNC8

like I said pure data. or maybe its all above my futile mind

 

[sberry]

Cheaper to buy enuf comp than to test it all and you are left with the comp in the end.

 

[sberry]

My Bud ran around trying to pay to test a 20 yr old fuel pump,,, I ask how much... 50$, I ask,,, well how much is a new fuel pump? 50$

 

[sberry]

That is a cool meter though.

 

[matt_i]

The biggest question is how much one wants to invest in flowmeters. The classic design is basically a "throttle" where there's a tapered neck in a tube, the pressure is measured accurately with "inch-of-water" class pressure levels on both sides of the throttle, and its calibrated somewhere else to give the end user a "lookup table" to find the flow based on the pressure differential. There are turbine flowmeters, sort of a fanblade in a tube, the RPM of the fanblade is sensed accurately and this translates into a flowrate. Easy to spend 2-5 thousand on such a setup.

The next biggest question is how to control standard conditions of the air. You might know about the ideal gas law, since air is compressible, it will do all sorts of things that affect the outcome of the flowmeter. Humidity, temperature, altitude, and the barometric pressure will all influence this. To control this accurately, theres a setup called "standard air machine" which is probably on the order of $25k-50k to buy. I'd SWAG easily 10-15% difference in results from precision flowmeters above, used without the standard air setup. This brings the "CFM" up to "SCFM"...

For a company mass producing the air compressor, the engineering costs are just part of the product development investment, to be recouped later thru manufacturing and selling the product. In other words, too expensive for an individual.

My involvement in this was working on the development of internal combustion engines. Many needs to measure air and liquid flow. Many needs to standardize atmospheric conditions for accurate apples-to-apples comparisons. I don't do this any more, so some of my information is probably dated. But recall those days fondly.

One can do the physics or math and figure that power in roughly equals power out, minus some losses. Measuring electrical "pressure and flow" is done with volt and ammeters, measuring the mechanical "pressure and flow" is done like above. The key reason that one can't just measure amps/current, multiply by a basic voltage and arrive at the result by conservation of energy is there's no telling who makes a more efficient compressor. In other words, getting more airflow and pressure for the electrical energy put in. The poor ones will generate more heat. The better ones, less heat.

 

[jonjon1]

So whats the point of all of this?

You buy a product from a reputable company, take good care of it, and spend your time using it NOT thinking about how to make sure it is usable....


I buy a compressor, fill it with royal purple oil, wire it (with an hour meter), plumb it, and start it up, time how long it takes to shut off, record that, the psi, the date, and the hours in my notebook. Then after 500 hours, I change the oil, check the belts, and filters, and time it from 0 to shut off again, if its about the same as it was new, then I write down the new info and go about my day, lol...

 

[Fixnair]

Boy did you open a can of worms. I have spent my life in the air compressor business. I am not an engineer but I can tell you that a reputable industrial manufacturer will have published volumetric efficiency curves and piston displacement. From this data you can calculate SCFM at any given pressure within the limits of the machine.
To test these claims they adhere to exact standards. These include barometric pressure, humidity altitude and temperature. To duplicate these standards in the field would be expensive to say the least. To test any other way would generate more questions than answers.
As far as wear affecting performance there is an easy test for a two stage compressor. Interstage pressure is directly related to capacity. It should be anywhere from 28 to 32 PSIG. If it is lower than that delivery is deminished. If it is higher than that delivery is not diminished much but power consumption goes up as well as heat.