Impeller Speed

[Chris Parks]

I could PM Michael Clark with this question but I thought it may interest others and perhaps there might be contributions from others so technically minded.

Is there an ideal speed for an impeller?

How do we know what that speed is?

I use my Clearvue 15" impeller as a reference for this question. Because it has a VFD I can run it at any speed I like and the speed at 60hz is only used because that is what the US cyclones run at. Has it reached its ideal running speed at that RPM? I have run it faster, up to 70hz and nothing blew up but I am guessing there is a point of diminishing returns and a point where it won't pay to run it any faster.

[Michael W. Clark]

Hi Chris,
As you increase fan speed, the HP requirement goes up by the cube, but the CFM only goes up linearly. Therefore, increasing speed is generally less energy efficient. I am sure there is a sweet spot, but it is most likely fan/manufacturer dependant. The fan manufacturer's I have worked with generally spec a fan so that the operating point lands on the curve so that it is percentage of the maximum SP. A good rule of thumb is 80-85% for industrial fans, but each manufacturer is different and high SP applications may be chosen differently.

Most AMCA fans have a maximum safe operating speed published, not sure about others. The belt drive industrial fans spec'd for a particular application generally operate very close to the maximum safe speed (within 5-10%).

Mike

[David Kumm]

Most impellers have an rpm rating so you need to be careful. I've played quite a bit with adjusting speed of various impellers and found most of the bang for the buck is within +- 5 hz. It will be dependent on your ducting and resistance so some extent. A BI fan will not benefit as much from a speed increase as a straight blade will as the BI are designed to operate within a narrow window and so they won't overamp a motor. In your situation i would guess that there will be some improvement in cfm but I would not increase hz very much without checking the impeller specs or whether the ODP motor can handle the additional heat. I would prefer a TEFC motor. You really need to measure the cfm to figure out the sweet spot of your system. When I ran an Oneida 15" I gained 100-200 cfm at 63 hz but after that I got little increase. My straight blade will gain 300-400 cfm at 65 hz but the design of the blades allows for more cfm under pressure and a BI. Dave

[Chris Parks]

So what we might agree on is the sweet spot is dependent on the total system design. Would it be a fair observation to say that we should measure the system performance at the exhaust as that is the sum total of everything. In fact it would be possible to measure different parts of the system like out of a TS or out of a bandsaw etc.

[David Kumm]

Chris, I measured at different machines, particularly the ones I was trying to juice up. Long story, I have a big 3 run 50+ ft per run 25 drop system and through brilliant planning kept upgrading machines and putting them at the end of the runs. My 24" Oliver planer and 48" Oakley edge sander were the worst offenders as old machines are great but their DC is horrible. I could have compensated somewhat by rerouting ductwork and going 8" all the way instead of necking to 7" but the cost of the fittings and gates- some are pneumatic- was more than a used blower and motor on ebay. I found a looked to never been used Cincinnati RBE 9 with a gold Baldor 735 hp motor for 700 delivered. Some acoustical blankets and a muffler cost another 300 or so but now I can pull 2000+ cfm through 7" mains into 6" ports so my DC is as good as the old machines can get. Prior to that i tried increasing a 15" Oneida but the design didn't allow for any more cfm than 10.5 amps would pull. I switched to a 15" straight blade off another blower and increased the cfm but at an amp cost of about 3 so the 5 hp motor was at the limit for only 200 cfm or less. The RBE runs a 15.75x6.25 staight blade and I run at 55 hz when just cleaning the air to 65 hz when running the sander. Most of the time 60 hz is just right and pulls 14-15 amps. at 65 the amp draw goes to 17+ so I max out the motor. the fan is rated for 4500 rpm. Dave

[Michael W. Clark]

So what we might agree on is the sweet spot is dependent on the total system design. Would it be a fair observation to say that we should measure the system performance at the exhaust as that is the sum total of everything. In fact it would be possible to measure different parts of the system like out of a TS or out of a bandsaw etc.

And I would go a step farther to say that the sweet spot is also dependant on the fan selection, both type and manufacturer as David said. For industrial DC systems, the system is designed first (starting with hoods, then ductwork) then you have all the information you need to select the fan. I know home/hobby DCs are not done this way, but certain assumptions are usually made that would have a system "work" for the majority of users.

As far as measurement, it depends on the purpose of the measurement. If you have a branch that is nortoriously poor DC or you know it is your branch with the most loss, you could measure at the branch and optimize from there. In this case, you should be better at all the other branches. If ths is a multi-branch system (as most industrials are) you have to consider balance between the branches. If you are trying to rate your collection equipment flow capability (cyclone/fan/after filter as applicable), then I would measure at the collection equipment inlet. This way you get the flow and can measure the corresponding system static pressure in the same hole. A flow reading means very little without the static pressure component.

Mike

[Chris Parks]

The question still remains, is my impeller turning at an optimum speed and how is that determined. Can I simply put a manometer in the inlet transition of the cyclone, ramp it up in equal increments and plot the results? What do I run the flow figures against, amps, RPM or both. I suspect amps will not have enough resolution to be useful.

[David Kumm]

I would open enough gates near the cyclone to max out the impeller and cfm. You will know that because the amp draw will increase with the gates until they hit whatever the system max is and top out. I used an anemometer at the open ports and they all will read basically the same. I then added or subtracted hz at 2-3 increments and recorded the cfm and amp draw. Others will chime in with ways to measure but since i was concerned with velocity or cfm at certain machines, it was easiest to just use my Velocitor and read the results. Dave

[John McClanahan]

Air has weight and the impeller blades throw the air. The faster the motor runs, the harder the blades try to throw the air, resulting in more suction. The problem is that induction motors can only run at 3600 RPMs at 60 cycles AC. That's where the diameter of the impeller gets comes into play. Shopvacs have small impellers but spin around 20,000 RPMs to make up for the size.

John

[David Kumm]

Air has weight and the impeller blades throw the air. The faster the motor runs, the harder the blades try to throw the air, resulting in more suction. The problem is that induction motors can only run at 3600 RPMs at 60 cycles AC. That's where the diameter of the impeller gets comes into play. Shopvacs have small impellers but spin around 20,000 RPMs to make up for the size.

John

The OP here has a three phase motor on a vfd so the speed and hz can be adjusted. Shop vacs pull low cfm but handle high pressure whereas a DC impeller is the opposite- high cfm but can't handle much pressure. The diameter helps with that, increasing the speed does somewhat as well but at a high energy cost and if running a curved blade the increase in speed might not result in much pressure or cfm increase depending on where you start on the system curve at 60 hz. Predicting any outcome is way above my pay grade so anyone doing this needs to measure the real results on their system. A system with a 5 hp motor, 8" mains, and a 14-15" impeller will see more improvement than someone with the same layout and a 16" impeller because both will top out at FLA, just at different speeds. Dave