Honda 1.8L running on test stand video

[charosenz]

I am sure my tanks are wider than yours so I definitely will be fairing them out of the flow. My tank will also be about 2/3 of the way "in" the box. Since my cowling will protrude down in the airstream a bit my opening overall will be larger. I have to make it wider too. I am leaning towards 4"H x about 18" wide for the inlet and about 5"H x 18" outlet. The outlet will not have a swing ramp to vary the outlet. If it cools too much I may consider adding an adjustable flap up front, but that will be put off for quite some time after much testing.

As far as attachment goes, I am lucky that I have "L" aluminum stiffners in the floor right above the front "walls" above the tank. I probably will use driven rivets there and pulled rivets the rest of the "L" angle where it attaches to the belly. All angle will be 3/4" 0.63" 6061. I plan to use #8 screws/nutplates for the side walls of the box where they attach to the angles at top. The rest of the box will be LP4-3 rivets from Vans.

The tank will be its own section 8" long. The back will be a section and the front another section all 0.32" 2024-t3 aluminum sheet from Vans. Over all it should be about 8.5" tall, 24" wide at the tank, and about 43" long...

Feel free to comment on the plan...

Pic of shorten cardboard mock up.....




Charlie

[DanH]

Originally Posted by charosenz

I am leaning towards 4"H x about 18" wide for the inlet and about 5"H x 18" outlet. The outlet will not have a swing ramp to vary the outlet. If it cools too much I may consider adding an adjustable flap up front, but that will be put off for quite some time after much testing.

Feel free to comment on the plan...

Those are very large areas, both inlet and outlet, in particular given the theoretical higher efficiency of a water to air exchanger. And closing the inlet in order to to throttle flow would be bass-ackwards. You want to throttle the exit. "Swing ramp" sounds right.

Before going too much further down this path, are you sure you can't move the exchanger(s) inside the cowl? Most sport airplanes get real slow with something the size of a garden wheelbarrow strapped to the belly. You're a good fabricator; treat it as a challenge. Heck, Eggenfellner did it with the Subaru packages, and I note SARL racers with nicely integrated water cooling.

[Jpm757]

Given any thought to a P-51 style scoop?

[charosenz]

Originally Posted by DanH

Those are very large areas, both inlet and outlet, in particular given the theoretical higher efficiency of a water to air exchanger. And closing the inlet in order to to throttle flow would be bass-ackwards. You want to throttle the exit. "Swing ramp" sounds right.

Before going too much further down this path, are you sure you can't move the exchanger(s) inside the cowl? Most sport airplanes get real slow with something the size of a garden wheelbarrow strapped to the belly. You're a good fabricator; treat it as a challenge. Heck, Eggenfellner did it with the Subaru packages, and I note SARL racers with nicely integrated water cooling.

Dan,

No doubt this will surely add drag to the ship. But honestly that scoop (while a aberration for an RV purist) is actually starting to grow on me...But yes, I have spent hours trying to squeeze it in the cowl and the only way that would be possible would put the radiator sideways to the flow and even then it would be real tight. I know a few of the rotary guys used condenser cores up front and other custom radiators but this engine is tight up front on the right side - just the way it turned out....But yes, I think with some custom work and probably using two small radiators you could pull it off. Or if you build a custom cowl, but I don't want to do that again....

Actually the other concept I seriously considered was laying a thin one inch core flat against the belly. This would greatly reduce the side view (Height?) of the scoop and help a lot in the Looks category. I think a one inch core that was about 24" x 18" would work quite well. There would be some challenges with the tubing, but nothing too tough. Ultimately I went this way because I know several have done it and had very good success with it.

The next post asked about a P-51 style scoop. Yes, absolutely I did consider that. If you mean just a small scoop in the belly with the radiator aft in the fuse..... The only downsides this would be what it does the balance and that far aft with this airframe and the internal ducting for outlet was enough to cause me to pass.....And of course you would still end up with a inlet down in the airstream under the belly so not a whole lot of "gain" and lots of extra work. But yes, I thought of it.

[BillL]

Originally Posted by charosenz

Dan,

No doubt this will surely add drag to the ship. But honestly that scoop (while a aberration for an RV purist) is actually starting to grow on me...But yes, I have spent hours trying to squeeze it in the cowl and the only way that would be possible would put the radiator sideways to the flow and even then it would be real tight. I know a few of the rotary guys used condenser cores up front and other custom radiators but this engine is tight up front on the right side - just the way it turned out....But yes, I think with some custom work and probably using two small radiators you could pull it off. Or if you build a custom cowl, but I don't want to do that again....

Actually the other concept I seriously considered was laying a thin one inch core flat against the belly. This would greatly reduce the side view (Height?) of the scoop and help a lot in the Looks category. I think a one inch core that was about 24" x 18" would work quite well. There would be some challenges with the tubing, but nothing too tough. Ultimately I went this way because I know several have done it and had very good success with it.

The next post asked about a P-51 style scoop. Yes, absolutely I did consider that. If you mean just a small scoop in the belly with the radiator aft in the fuse..... The only downsides this would be what it does the balance and that far aft with this airframe and the internal ducting for outlet was enough to cause me to pass.....And of course you would still end up with a inlet down in the airstream under the belly so not a whole lot of "gain" and lots of extra work. But yes, I thought of it.

Ross did some extensive testing on his cooling, I'll look it up, but if memory serves, he had quite low drag. High exit velocity. The heating of the air makes up (some) for the drag across HX. It would be wise to simply improve on his proven work.

[DanH]

Originally Posted by BillL

Ross did some extensive testing on his cooling, I'll look it up, but if memory serves, he had quite low drag. High exit velocity.

Quite low cooling drag (internal flow)...an attempt to prove the Meredeth Effect, i.e. thrust created by an exit velocity higher than freestream. And he did prove it, sort of. The published maximum exit velocity value (from Kitplanes) was 104% of freestream at 80 KIAS.

That's very good, but note that at 80 knots, the actual freestream velocity delivered to the diffuser inlet is much higher than indicated airspeed, due to the huge propeller outflow component. The prop outflow effect can also be seen in measurements of an aircooled installation:

http://www.vansairforce.com/communit...&postcount=198

Even if the exit velocity was less than 100% of freestream at cruise speeds (and I suspect it was), it is still very likely a lot higher than the average Lycoming cowl. That part is a win, a reduction in drag due to internal flows.

Returning to the context of this thread, Ross wrote that he made no attempt to measure the external aerodynamic drag of the installed belly pod. Further, the system as proposed retains flow through the cowl, another drag cost. I'm suggesting Charlie consider all his options before hanging a wheelbarrow-sized pod on the belly. If the exchanger(s) can't be fitted inside the cowl, then perhaps take a look at the beautiful underwing ducted exchangers we recently saw on an RV-8 in Oz. Simply moving the pod out of the propeller outflow would mean a significant reduction in external drag.

[rv6ejguy]

I picked up a solid 5-6 knots with my belly scoop over nothing there before (well some other small inlet and outlet ducts feeding from multiple HXs). When viewed from the front, there isn't much frontal area added since the stock exit duct area covers most of the rad duct frontal area. Looks like the momentum recovery offsets most or all of the duct drag in my case.

Cowling mounted can work but the result is more weight forward, poor momentum recovery and often marginal cooling.

I've followed numerous rad layouts and had internally mounted rads on my own plane for years. I spent a lot of time changing it all over to the present layout and couldn't be happier. It's lighter, faster and improved the C of G while cooling much better than previously.

You shouldn't need more than 25 square inches of inlet area with a proper duct shape to cool 180hp in flight. Extra area may be beneficial in ground cooling. With extra inlet area and a movable exit door, your cruise drag penalty is probably slight. Lots of compromises involved here. Certainly you don't want to overheat on an 80-90 degree day with a long ground hold somewhere.

I wouldn't look to Eggenfellner's layout as a good model. Lots of folks had marginal or inadequate climb cooling and had to step climb or cruise climb on hot days.

[rv6ejguy]

Originally Posted by DanH

Quite low cooling drag (internal flow)...an attempt to prove the Meredeth Effect, i.e. thrust created by an exit velocity higher than freestream. And he did prove it, sort of. The published maximum exit velocity value (from Kitplanes) was 104% of freestream at 80 KIAS.

That's very good, but note that at 80 knots, the actual freestream velocity delivered to the diffuser inlet is much higher than indicated airspeed, due to the huge propeller outflow component. The prop outflow effect can also be seen in measurements of an aircooled installation:

http://www.vansairforce.com/communit...&postcount=198

Even if the exit velocity was less than 100% of freestream at cruise speeds (and I suspect it was), it is still very likely a lot higher than the average Lycoming cowl. That part is a win, a reduction in drag due to internal flows.

Returning to the context of this thread, Ross wrote that he made no attempt to measure the external aerodynamic drag of the installed belly pod. Further, the system as proposed retains flow through the cowl, another drag cost. I'm suggesting Charlie consider all his options before hanging a wheelbarrow-sized pod on the belly. If the exchanger(s) can't be fitted inside the cowl, then perhaps take a look at the beautiful underwing ducted exchangers we recently saw on an RV-8 in Oz. Simply moving the pod out of the propeller outflow would mean a significant reduction in external drag.

It's good to consider all options but we who've already done it, have.

Most of my flight testing was done at 100-120 KIAS at around 6000 MSL. Yes, I had no way to quantify where the 5-6 extra knots came from- the magic properties of momentum recovery or the removal of several other ram ducts, NACA ducts and exit ducts.

Prop outflow is significant within a foot or so of the prop and drops off rapidly with distance. You can see a lot of the F1 guys run the carb snorkel within a few inches of the prop and see some very useful gains in MAP. My inlet is almost 40 inches from the prop disc. I did a test at some point to try to determine this figure but not sure if I recorded it. I'll consult my old notes. I only recall it wasn't very significant on my installation in flight.

Several Unlimited P51s experimented with removing the rad duct over the years and found little to no gain in speed. This again suggests that, properly done, a good duct has minimal, if any penalty. They could have fitted wing rads if they were better but other tests on the Spitfire and BF109 post war, showed that the short duct has high inlet separation which means more drag ultimately. No surprise there, though they could have been vastly improved with a guide vane I'd think, judging from what Russell Sherwood and I both saw in testing.

The big minus is that rad ducts on the wings outside the prop arc guarantee overheating on the ground within minutes on a hot day. Simply not practical for a GA airplane.

[DanH]

Originally Posted by rv6ejguy

Prop outflow is significant within a foot or so of the prop and drops off rapidly with distance.

Not that simple. Consider an RV at 1800 RPM on the runup pad. Airspeed is zero, yet prop outflow would make it hard to stand upright behind the tail. Stand behind one at 2700 RPM, tied to a truck, if you can.

Prop outflow as an incremental addition to freestream drops off with increasing airframe velocity, or decreased power.

My inlet is almost 40 inches from the prop disc. I did a test at some point to try to determine this figure but not sure if I recorded it. I'll consult my old notes. I only recall it wasn't very significant on my installation in flight.

It would not be at cruise settings. Was the 104% at 80 measured with climb power, level at steady low power, or with engine at idle, in a glide?

The big minus is that rad ducts on the wings outside the prop arc guarantee overheating on the ground within minutes on a hot day. Simply not practical for a GA airplane.

Interesting point. Was Geoff Braddock having that problem?

http://www.vansairforce.com/communit...d.php?t=145250

BTW...

Originally Posted by rv6ejguy

This is pretty much the recipe I'd use today if I was doing another RV using an alternative engine.

[rv6ejguy]

Originally Posted by DanH

Not that simple. Consider an RV at 1800 RPM on the runup pad. Airspeed is zero, yet prop outflow would make it hard to stand upright behind the tail. Stand behind one at 2700 RPM, tied to a truck, if you can.

Prop outflow as an incremental addition to freestream drops off with increasing airframe velocity, or decreased power.

It would not be at cruise settings. Was the 104% at 80 measured with climb power, level at steady low power, or with engine at idle, in a glide?

Interesting point. Was Geoff Braddock having that problem?

http://www.vansairforce.com/communit...d.php?t=145250

BTW...

Indeed, full throttle on the ground on my plane produces a couple inches of pressure at the cowl cheeks, no dispute there. In flight, a different story. Confirmation by measurement. If I get a chance, I'll instrument again and fly if I can't find my old data.

Now, if we look at something like Strega which has had hundreds of mods done to make it faster, including a massively revised scoop, radiator and cooling system trying to find every knot through drag reduction, you think they wouldn't have gone to wing mounted rads if they thought they would have gone faster?

My data was taken at steady state, straight and level, stabilizing for a couple minutes in most cases. Trying to measure for my purposes in the climb would have thrown more variables into the mix.

Looks like Geoff's installation would have some of the rad face in the prop blast. Certainly if rads our mounted outboard of this, you'd only have still air conduction on the ground and essentially zero cooling. No delta, no flow, no cooling.