Measuring Solar Collector Airflow

This may be a wild idea, but I would consider using a regenerative blower discharging through a small pipe, where velocity would be enough to measure, followed by perforated metal distribution plate matched to the shape of the collector you are evaluating. The velocity in the collector would be velocity in pipe × cross-section of pipe, ÷ cross-section of collector. A valve in the pipe could adjust the air flow.

I didn't know what a regenerative blower is, so I looked it up. It took more effort than I expected. Here's the first definition I found:

from www.egl-neon.com/glossary.html:

A type of blower very suitable for a neon shops air supply needs.

Here is a more useful one, from Pneumatics Online: Regenerative Blower:

Regenerative blowers in plant installations are typically in a direct drive design, versus electric motor / gas engine belt drive configurations. (refer photo depicting these 2 designs). The impeller in the direct drive (sometimes called monobloc) construction is mounted directly on the electric motor shaft and rotates at the motor's nominal speed, typically 2900 or 3500 rpm. The impeller consists of numerous radial blades on the circumference of the impeller. The number, size and angle of these blades contribute to the pneumatic performance (flow vs. pressure / vacuum) characteristics of each blower. Some makes of blowers have somewhat "flat" curves (Y axis =pressure / vacuum, X axis = scfm), while other makes have quite steep performance curves. The impeller spins within a housing that consists of an inboard and outboard "channel" (hence the term side channel blower). As the impeller passes the inlet port, air is drawn in. As the impeller rotates, air is captured between each blade on the impeller and is pushed both outward and forward into the channels. The air then returns to the base of the blade. This process is repeated over and over as the impeller spins. It is this regeneration that gives the blower its pressure / vacuum capabilities. In essence, a regenerative blower operates like a staged reciprocal compressor and while each blade to blade regeneration "stage" results in only slight pressure increases, the sum total, from air entry to outlet can yield, in some makes continuous operating pressures up to 9 psig or vacuum to 14" hg with flows in the 200 to 250 scfm range at these points.

Most blowers are single stage as air molecules travel around the blower housing one time and then are exhausted. In addition to single stage units, some blower manufacturers offer two staged blowers. Two stage regenerative blowers are capable of providing nearly twice the pressure or vacuum of single stage units. Two staged blowers operate much like single stage units as the impeller strikes air molecules over and over to create vacuum and pressure. In a staged blower air molecules will make one revolution around the front side of the impeller. Instead of being exhausted after the first revolution (like single stage units), the air flow is channeled to the back side of the impeller through internal porting. Air molecules will then make another revolution around the back side of the impeller thus doubling the number of times that the impeller blades strike the air molecules. After a second revolution around the blower housing, the air is exhausted. Higher pressures and vacuums are able to be generated with two stage blowers since the impeller blades strike or "regenerate" air molecules through two revolutions instead of one.

And a caveat from Regenerative Blower:

Caution: Regenerative blowers operate with very tight tolerances. The rotor runs at a very high velocity. Ergo, the unit tolerates very little abuse. If you decide to disassemble your blower, do not beat or pry on the housing if you expect the rotor to not drag afterward. Even a tiny opening in rotor-diffuser clearance will cause a dramatic decrease in performance.

You used an incline manometer with the pitot tube?

Hi,

"You used an incline manometer with the pitot tube?"

Yes, I have used it and also have a Dwyer Magneheilic gage that has a full scale range of 0.2 inches of water. Both work fairly well -- the pressure I am reading is typically about 0.15 inches of water.

I did not mention it, but adding the smaller diameter (3 inch) section of duct for the pitot tube measurment tends to back presssure the collectors and cause the glazing to bulge -- this is managable but it would be nice to have a good measuring technique that did not do this.

From the point of view of getting stable and consistent readings, I'd like to make the 3 inch section even longer (its about 12 diameters now), but this would cause more pressure drop, reduce flow and increase the back pressure.
Moving the point at which the pitot tube reading is taken a couple diameters upstream or downstream from where I normally take it does effect the reading, so, it seems like a longer run of 3 inch duct would be desirable.

Gary

With this browser I cannot copy and paste text so that explanations follow witha another Sorry for 2 comments in stead of one but such are the limitations.

Herer are the explanations. A simple design which can be manufactured by almost everybody can be as follows. It consists of a glass tube whose diameter can be defined according to the flow domain and a "swimmer" which slides in the tube with a small gap. The swimmer let air flow around through machined windows. The force balance is : Axial component Fa= G*sin (α) Radial component Fr=G*cos(α) Friction at the wall level due to sliding Ff= Fr*µ The air flowing around the swimmer generates a force which is Fp= K*Q²*Aswimmer The swimmer is in balance when Fp= Fa± Ff depending on the direction of sliding. The final equation for the flow will be: Q²= G*(sin(α)±μ*cos(α))/(K*Asw) = [G/(K*Asw)]* (sin(α)±μ*cos(α)) = C*(sin(α)±μ*cos(α)) C is a constant since it contains values which one for ever defined at assembly. It is thus only needed to change the angle and see at which position the swimmer is not moving. The friction effect can be eliminated by coming once from high angles to small and a second time from small to big. The true angle will be the average value of the 2 obtained values (in the assumption that μ is constant). The procedure is simple one starts for instance with a small angle and the flow force being > the one given by the weight the swimmer goes to the left end. The angle is incremented till the swimmer comes in the middle (about) of the glass region. This angle is put down. The angle is increased further till the swimmer goes right and then decreased slowly till again an about in the middle position is obtained and a stable swimmer. This is the second value. If the nominal dimensions are chosen so that the working range is between (20)30 and 60(70)° the resolution can be OK. There is another approach to eliminate the friction effect but it is more difficult to obtain. It is enough to make the air flow such that the swimmer rotates. It is known the friction in normal direction will be zero. I can explain why but it is not the place to do it. Once calibrated the design can be scaled up and down since some coefficients are only geometry dependent (relative dimensions). For a low friction the swimmer can have a central body of SS 304 or brass and the other parts of PTFE.

It looks as if you have tried most of my ideas already. How about using a positive displacement pump to produce a steady output? You don't need the pressure they are designed for, but they should meter out the air pretty evenly at constant near-zero pressure. A vane pump is probably the least overbuilt option, but piston pumps would do as well. A home-made bellows might work, too.

If both arrays are in the same air stream, they experience the same air flow.

Construct a duct with a glazed top to admit the solar light and wide enough to fit both arrays side by side. Feed the output from both fans into a plenum chamber and connect the plenum to the duct with one or two full width slots. Leave enough room between the slots and the arrays to create laminar flow if you require it. Constrict the exhaust from the duct to 1% of the duct cross section. You can now measure the airflow within the exhaust restriction with 2 orders of magnitude more accuracy. The air within the duct will be positive with respect to ambient but this should have no effect on your comparisons.

Re: If both arrays are in the same air stream, they experience the same air flow.

I'm confused. What is implied in your first sentence is that if they are both *in series* in the same air stream, the flow is the same. What you then go on to describe seems to be an arrangement with them in parallel?

Sorry don't have time for a 1000 words, a drawing will have to do

Rough plan (slots would be horizontal across whole width, not vertical as shown)

Thanks for the response! I'm an EE, and I guess I don't really understand laminar flow very well. If you're convinced, I'll just take your word for it (at least until I feel like digging into laminar flow).

Just for the record, note the misunderstanding mentioned in post #14.

Hi JH,

Thanks for the idea.

I guess I am not seeing why the flow through the two collectors is going to be the same? They are of different designs and have different internal resistances to airflow, so feeding them both from the same constat pressure plenum does not seem (to me) to make the flows through each the same? Maybe I'm missing something?

Gary

PS thanks to all for the responses -- think about all of them. The vane positive displacement vane type fan would be a nice simple solution if I can find them at a reasonable price.

My mistake. I assumed that you wanted to test an identical flow past the collectors, but you actually want to test an identical flow rate through the collectors.

You want to seal off all but the inlet and force all the flow inside. Measuring flow in a restricted exhaust will still give you a multiplier that will provide you with much more accuracy. I don't follow why you need to test the units in parallel. Input the measured flow rate into a PID loop and use it to adjust a damper or a variable speed drive on the fan. Adjusting the set point on the PID allows you to test at all known flows, and it is reproducible so you can test units sequentially with exactly matching conditions. Because everything except the collector under test is then common, all the other variables that could affect your result will cancel out.

Hi,

My basic problem is that the two collectors have different internal resistances, and I need to establish exactly the same flow rate through each collector.

The heat output from each collector is proportional to (temperature rise)*(flow rate), so if the flow rates are the same, then the more efficient design is the one with the highest temperature rise. The basic problem I am having is that I'm finding it difficult to come up with a good and consistent way to measure the flow rate.

In effect, I am already using the exhaust restricition idea by taking the exhaust flow down to a 3 inch diameter duct to get the velocity high enough so I can measure with a pitot tube.

Gary

I think you are describing two plenums in paralell? If so, then a simple tube running between them with a glass chamber in the middle should do. In the glass chamber all you need is a leaf suspended. If there is a difference between the two airsteams, then there will be a flow through the chamber and the leaf will tell you from which side it came. If all you need to know is the difference, the leaf will tell you everything you need to know.