Harmon Rocket II

Kevin Horton said:

...Unless I've screwed up this quick calculation, it seems unlikely that pressure from the prop can explain any more than 0.1 in HG of the MP rise we're seeing...

Click to expand...

I'll not comment on your math, but I can tell you that I've heard of this "prop boost" phenomenon many times over the years. I've told this story before, but on my Hiperbipe I fabricated a divergent duct inlet system which featured an easily adjustable pitot inlet. It was simply a 3" diameter aluminum tube which could be adjusted for length by sliding in and out of the plenum after loosening a single hose clamp. I could fly, take measurements, land, adjust the length, and be back at the test altitude again in 10 minutes. Long story short, the difference between frighteningly close to the TE of the prop and just slightly proud of the cowl OML (about 5 inches difference) was worth 0.3 of MP. Not very scientific, but the test was repeatable and consistent. Closer to the prop is better. One caveat: My inlet was about a foot outboard of the crankshaft centerline, so it was well outside the root area of the blade.

Well I kinda opened a can o' worms with them pics, didn't I .

I oughta get out there and do a constant DA (PA) test at various power settings, just to see how MP reacts to increasing speed.

Then perhaps a test at various altitudes at constant power to measure MP.

Kevin, to make such a test valid, I would assume using WOT on the first test would be valid, and just increase RPM?concur? On the second test, what would you recommend, keeping WOT and consistent RPM at the various altitudes, or should power setting somehow be equalized (change RPM?) to compare MP at various altitudes?

For your calculation assumptions, my OAT probe is on the right side of the fuselage, under the H-stab leading edge. I have measured a ram rise of 5 degrees in a constant altitude, varying speed test (slowed from 170 to 70, saw a drop of 5 deg, accelerated back to 170, saw the 5 deg rise). I figure this impacts my TAS readout, and have seen a consistent 2-4 knot TAS error (fast) during my speed tests, as indicated by the NTPS spreadsheet outputs. Not sure how this affects the MP readout.

My engine's dyno showed 322 HP (builder's dyno YMMV) in 1998. 1400 hours later, its still strong, but I don't think I've got the 330 ponies you assumed?I wish! (Next motor?want more?its a disease! )

Adding the James inlet did push the inlet closer to the prop, but I did not see a dramatic increase in MP, looking back at the spreadsheets.

How does one test the cal of the MP sensor?it, and the VM1000 are also 16 years, 1400 hours new. I'm game!

Fun discussion?thanks for the data and educ-ma-cation!

We've forgotten that we are talking about Rocket-Power tho, right! :roll eyes:

Cheers,
Bob

Easy to check the propeller's contribution to MP. Just rig two temporary pitot tubes, one in the prop outflow, the other in freestream. Run a line from each to the cockpit, set a steady flight speed, connect a manometer to each in turn. EDIT: See next post. Outflow ain't uniform, so you would also need to pick a sensible radius from crank center.

I remembered seeing data from a prop pressure rake somewhere. Found it in CR3405. Added definition is from the paper:



Couple interesting things here.

Obviously the pressure is not uniform. In close to the spinner showed very poor Cp, so a smiley inlet which hugs the spinner may not be a good choice unless your prop has excellent airfoil sections in the blade roots. EDIT: Retract that statement. In close to the spinner is showing 0.8, which is higher than freestream static by a good margin. It's just not picking up anything extra from the prop.

Assume freestream dynamic is 1.45 (@10,500, 210 KTAS, 51F), freestream static is 20.185, and the inlet location at a radius similar to cowl width (much like a Vans carb airbox snout). The plot seems to suggest maybe Cp = 1.04 at cruise. That works out to be 0.06" Hg (way to go Kevin!).

Given the same conditions, best possible works out to be 0.29 Hg at full power, unless you can figure out how to locate your inlet about 24" from the crank center.

The boost would be higher at a lower altitude, and a different prop could result in a big change.

When I redesigned my lower cowling a few years ago moving the engine intake air away from the spinner and closer to the prop was one of my goals. Paul Lipps was one of the inspirations for this change. I had asked Paul how large the inlet needed to be. (He immediately used one of the formulas that Dan Horton quoted in this thread, by memory, and in his head calculated the answer in a few seconds. I checked the math at home with a calculator and he was bang on, amazing.)
It was moved down and as far away from the spinner as I could while still maintaining a smooth line with the cowling. It extends forward until it is within 1/4" of the prop blade when the prop is in the full coarse setting.
With the exception of a prop change I could not demonstrate speed improvements from any one of my modifications, this one included.

Tom Martin said:

Paul Lipps was one of the inspirations for this change. I had asked Paul how large the inlet needed to be.

Click to expand...

Darn....I really liked Paul. RIP

Ok, I'm guessing he calculated a minimum practical diameter. I don't believe there is a perfect diameter.

Play with the stuff in post 48, substituting various intake rings. If you pick a full power climb speed, you can calculate the intake diameter which neither chokes the intake, nor provides any slowing of intake velocity, i.e. no MP increase due to dynamic pressure. For a 540 at 2700 and 100 knots, it's about 2.8" diameter. Below 100 knots it would choke the intake, and above 100 it starts adding ram. At 240 knots it would add about 142 knots worth of ram pressure. The actual MP boost would depend on the air density (altitude and temperature) at which the 240 knots is being flown.

A larger intake lowers the zero gain speed and increases pressure at maximum speed. Tom, if your Bower intake is 3.5", the zero gain speed appears to be around 63 knots, with about 177 knots ram at 240.

A designer can keep pushing the intake ring larger and larger. The above trend continues, but with with a steadily diminishing gain. Meanwhile external drag is probably rising.

So what is the perfect compromise? I dunno. I think in the end you just pick for what you want to optimize. I wanted climb performance above all (and I was unsure about inlet separation with a short airbox), so I went with 4". With the 390, zero gain speed is about 35 knots, meaning about 75 knots of ram at a typical 110 knot climb speed, plus whatever pressure I'm harvesting off the prop. Although the external diffusion itself is frictionless, it means I have more skin area, a blunter front end, and a possible external separation drag if the external shape is poor. These things may be costing me at high speed.

All the above assumes external diffusion, i.e. a low velocity ratio inlet. It all goes out the window with a high Vi/Vo inlet and internal diffusion, assuming you have the necessary cowl length for it. I don't, so I have not thought about it a lot.

Postscript: I put together an Excel spreadsheet for the above. There is a link to an online aero calculator if you want to get actual ram pressure for a particular altitude and temperature, without getting confused in the units.

Download here: https://onedrive.live.com/redir?resid=EC4690816EDB7227!108

Use at your own risk.

Dan,

Interesting stuff. The James inlet ring is actually 3" (9.42 in^2). Not sure how to enter that into the calcs shown in post 48. How would that impact the theoretical MP you calculated? (Couldn't tell where to plug that in on that page).

IIRC, your inlet lip radius is nicely formed, so separation drag should be low. Is your concern about higher speed loss due to the larger 4" radius inlet. Just trying to understand all aspects.

Cheers,
Bob

rvmills said:

The James inlet ring is actually 3" (9.42 in^2)..... How would that impact the theoretical MP you calculated?

Click to expand...

Compared to a 4" ring, it reduces MP by 0.4" Hg, given the same conditions as in post 48.

Is your concern about higher speed loss due to the larger 4" radius inlet.

Click to expand...

Umm....4" diameter

Just trying to understand all aspects.

Click to expand...

You and me both brother. Quantifying external inlet body drag in lbs would require CFD way above my lowly pay grade. Even then, it would be based on theory and input (remember GIGO?), and probably wouldn't be as good as an accurate measurement.

BTW, recall the data I sent last week? Rod Bower has some additional info on his website you might find interesting.



At zero ram, the loss for the 3" ring, connections, and open butterfly in the Bower airbox appears to be 29.29 - 29.10 = 0.19" Hg on the IO-580. Should be a little less on your 540. In any case, it's one more in a list of losses that reduce the pressure delivered at the intake valves.

Dan,

I'll need to dig into the numbers on 48 more in depth, thought you were saying that smaller was more efficient at higher speed, and larger was better at lower speed. Surprised the 3" dropped MP at speed.

Radius, Diameter...picking nits now, eh! (my bad, good catch).

Concur on the difficulty of testing the efficiency of cowling inlet lips.

Interesting data on the Bower site. I'll assume (uh-oh) that the velocity stack referred to is just the servo itself (which has a radiused lip, reminicent of a velocity stack), attached to the engine, and pulling ambient air in directly...no feed duct, minimal loss. Add a slightly divergent duct and a butterfly valve, losses increase...makes sense.

So perhaps my MP numbers appear high...will test more, see what is repeatable, and look into MP calibration.

As a side note, have had some discussions with Tom McNerney (Lancair guy who posts here) and Steve Smith about adding a true velocity stack to the mouth of the servo, inside a large divergent duct, to build pressure in the duct (plenum), then use the v-stack to minimize losses into the servo. Mixed opinions on it...but a possible downstream experiment, though at our pay grade, the quantification would be MP results, which as we've seen, can be pretty variable, due to environmentals.. This would be a filterless set-up, just for race mode.

Just fun ideas for discussion!

Cheers,
Bob

rvmills said:

... thought you were saying that smaller was more efficient at higher speed...

Click to expand...

Possibly, maybe, in terms of external cowl drag.

... and larger was better at lower speed.

Click to expand...

In terms of increased manifold pressure, the larger inlet is better at any speed.

Interesting data on the Bower site. I'll assume (uh-oh) that the velocity stack referred to is just the servo itself (which has a radiused lip, reminicent of a velocity stack), attached to the engine, and pulling ambient air in directly...no feed duct, minimal loss.

Click to expand...

"Large AFP Velocity Stack" probably means a machined test piece with a specified set of dimensions (to keep everybody equal in their tests), made for each brand/size servo. Here's one on an FM200, and you can see stacks for other servos in the background:

As a side note, have had some discussions with Tom McNerney (Lancair guy who posts here) and Steve Smith about adding a true velocity stack to the mouth of the servo...

Click to expand...

Have one molded into my airbox. Here is the original mold form, with machined nylon added for good detail in the velocity stack area:

Ah, roger all...good pics!

And any pics of the v-stack inside the airbox? Very cool! Shoulda known you were two steps ahead.

Maybe Tom will post a pic of his...might be a good topic for a new thread, since we are way OT here...but good discussion nonetheless!

Cheers,
Bob