Power from fuel flow - looking for data
 

The classical engine power charts provide power as a function of rpm, MP and altitude, but they are only valid if the mixture is set for best power. Many people like to cruise with some other mixture setting, so there is interest in a means to determine power either richer or leaner than best power mixture.

I've got an old Lycoming document that describes a method to determine power during cruise performance testing, using fuel flow as the main input. It isn't usable real-time in flight, but it could be used in conjunction with flight testing to produce power setting tables for a handful of combinations of rpm, MP, altitude and fuel flow. I've played around with this power calculation method a bit, trying to figure out if it produced consistent, credible results. If it does, I'll document it in a Kitplanes article and produce a spreadsheet to use it. If it doesn't provide consistent results, then I'll drop it. I'll give an overview of this old Lycoming method in the next message in this thread.

The big question I need to answer is does this method produce a consistent power no matter what mixture is used. I will answer that question by plotting speed vs calculated power, with points at the same rpm, MP and altitude, but at several different fuel flows. I've got a bit of data from my RV-8, but it will be some time before I can get any more, and I am not completely satisfied with the stability of my fuel flow indication, which makes is difficult to draw solid conclusions from my data. Also, I really need data from more than one aircraft, with more than one pilot, to see whether this method is useable in the real world.

I'm looking for a few intrepid RVators to help me test this fuel flow from power method. I need data from several aircraft with a Lycoming or Lycoming clone engine. The compression ratio must 6.75, 7.0, 7.2, 7.3, 8.0, 8.5, 8.7, 9.0 or 10.0 (one of the steps in the method is to look up iSFC, and the chart only has data for compression ratios that Lycoming sold). The engine displacement must be 235, 320, 360, 480, 540 or 720 (sorry, this document predates the 390s by several decades, and I assume the 290 was out of production when the document was produced). The aircraft must have a constant speed prop and a fuel flow system that gives current fuel flow and fuel remaining. I'm interested in data from engines with two mags, one mag + one EI, and two EI.

The test procedure is as follows:

Before flight, ensure you know the aircraft weight by accounting for all the stuff in the cockpit and baggage areas. If the item wasn't on the aircraft when the weight and balance was done, either remove it or figure out how much it weighs and account for in the gross weight. Fill the fuel tanks, and record the calculated gross weight with full fuel, all occupants, etc.

Find a test area with very smooth air, and no vertical air motion. No mountain waves, etc. Record the pressure altitude (i.e. with altimeter set to 29.92) and OAT. I don't care what altitude you use, as long as the air is smooth and has no vertical motion.

Set the desired rpm and MP, and don't change them for the duration of the test. Record the rpm.

Slowly adjust the mixture to find peak EGT, and record the fuel flow at peak EGT. Note: the test method assumes that all cylinders peak at the same fuel flow, but the real world doesn't work like that. Ideally you would record the fuel flow when each cylinder peaked (i.e. record four or six fuel flow values). Or, if they all peak at about the same fuel flow, give me an eyeball average of the fuel flow at peak EGT for all cylinders. Don't send just the fuel flow when the first cylinder peaks.

When looking for fuel flow at peak EGT, be very aware of how quickly or slowly your EGT system responds to changes in mixture. If your EGT system is slow to respond, you'll need to be very patient when adjusting the mixture, to let the EGT stabilize after each change. Otherwise it is quite possible to record a too low fuel flow at peak EGT.

Without changing rpm, MP or altitude, record level flight IAS vs fuel flow for a wide range of fuel flow values, both ROP and LOP. The wider the range of fuel flows the better, as long as the engine is running smoothly, with no misfiring. At each fuel flow, wait long enough for the IAS to stabilize, which may take several minutes. Record fuel flow, fuel remaining, and IAS.

For extra brownie points, you could repeat the above at one or more different conditions. I.e. change one or more of the altitude, rpm or MP. Data from several different flights is OK, as long as the aircraft CG remains pretty much the same - i.e. if you have an RV-4 or -8, don't do some flights with a passenger, and some without, as that will affect the relationship between power and speed.

Send me the following data:

Engine model
Engine compression ratio
Type of ignition system
Prop model (I don't need a detailed model number - I just want to confirm it is a constant speed prop)
Aircraft gross weight on the day of the test with full fuel.

For each altitude, rpm and MP that you have data, send me:
Altitude
OAT
RPM
Fuel flow at peak EGT for each cylinder

Then, for each mixture setting, send:
fuel flow
fuel remaining
rpm
altitude
OAT
IAS
Remarks - I am particularly interested in the stability of your fuel flow indication. I.e., with constant rpm, MP, altitude, mixture control, how much does the fuel flow indication vary up and down? Knowing this will help me interpret any noise in the results - i.e. is the noise due to issues with the method to calculate engine power, or is it possibly due to uncertainty in the fuel flow data).

I created a test card in Excel (and also PDF format) - Updated test cards uploaded on 25 May 2009 at 1715 EDT. I also created fixed pitch prop test cards in Excel (and also PDF format).

Caution - it isn't smart to run at peak EGT or lean of peak at too high a power. Use your best judgement on what power settings to use for these tests. Be nice to your engine.

The method to determine power from fuel flow is fairly interesting. It starts off with the fuel flow at peak EGT. Lycoming claims that the fuel flow at peak power is 1.178 times the fuel flow at peak EGT. You multiply the fuel flow at peak EGT to get the fuel flow at peak power. You take the compression ratio, and look up the indicated specific fuel consumption (iSFC) at best power mixture. Multiply the calculated fuel flow at peak power times the iSFC, to get the indicated horsepower (ihp) if the mixture had been set to best power. Indicated power is the power produced by the pressures in the cylinder, before the loss due to friction.

The method provides curves that show how the power varies with fuel flow both rich and lean of best power mixture. You take the ratio of actual fuel flow to calculated fuel flow at best power, and go into these curves to determine the indicated power at the actual fuel flow. Next, you go into another set of curves that give you the friction power as a function of rpm, engine displacement and whether it is a geared or ungeared engine. Subtract the friction power from the ihp to get bhp - the power that is available at the crankshaft to spin the prop.

The concept behind this method looks sound. But I don't know if the actual curves are correct. I also don't know how well it will work with engines with electronic ignition, as that will affect the iSFC somewhat.

The curves of friction power are interesting. They highlight the point that if you want to produce a certain amount of power, you are better off doing it at low rpm and high MP than at high rpm and low MP. For example, with an O-360 or IO-360, 22 hp are lost to friction to turn the engine at 2300 rpm. If we spin the engine at 2500 rpm, the friction power is 25.9 hp. So, if the power we want can be made at either 2300 or 2500 rpm, if we choose 2300 rpm we'll get 3.9 more horsepower for the same fuel flow than if we used 2500 rpm. Or, if we kept the power the same, the fuel flow at 2300 rpm would be about 0.25 gal/h less than the fuel flow at 2500 rpm.

Kevin,
I think Scott, Larry and I can probably accommodate you with some very comparative data( See oil cooler thread). We like having a mission.;)

Fuel flow, power and true airspeed
 

Kevin,

Man, you are going after data big time on this subject. Typical engineering effort. :) My take on it is much simpler.

I have always use fuel flow relative to a given TAS for flight planning purposes in these airplanes. It's the only thing that makes consistent sense to me. It takes a few data gathering flights to build a simple chart based on these factors. With the Subby, this information was developed up to 12,000' and included a still air MPG factor. That made it possible to pick an efficiency level appropriate for the flight, set the fuel flow and let it roll.

Consistent leaning technique is a given. That was not an issue with Subby but is with Lycoming. I used the fuel flow method of setting power in a previous Lycoming powered airplane with a fixed pitch prop and found it to be quite accurate for flight planning purposes.

I agree that if you have established a very consistent leaning technique, then looking at TAS vs fuel flow as the prime cruise performance test data is a good option. You don't need to do a huge amount of testing, as there are ways to gather TAS vs fuel flow data at one weight, altitude and temperature, and correct that data to predict performance at other weights, altitudes and temperatures. I'll eventually submit an article to Kitplanes that covers this.

But, there many good reasons why people are interested in knowing engine power, so I am looking at this old Lycoming method to see whether it will provide a useful answer to that question.

I like your plan Kevin - if you get enough data, you should see interesting results! One thing that you might watch out for though is the variation in fuel flow measurements based on where/how people have mounted their fuel flow 'ducers. I routinely see a small, periodic variation in FF during cruise - a period of about 10 seconds, variation of about 0.3 gph. So folks will have to patiently do a little "eyeball averaging" before writing down a data point.

Oh yes - smooth air, consistent atmosphere....reminds me of the Shuttle Crosswind DTO (Detailed Test Objective) we defined back in 1988. We wanted a few data points at a couple of different crosswind conditions. Because "we get what we get" on our few landings each year, we are still looking for a few data points....;)

Paul

Yeah, the quality of the fuel flow data is one of my big concerns too. I wonder how much of that 0.3 gph variation is due to the float valve in the carb modulating the fuel flow, and how much of it is due to inaccuracies in the fuel flow system itself. My fuel flow indication moves around a lot more than I would like, and it may limit the usefulness of this approach in my aircraft. The one thing that we do have in our favour is that the calculation is much less sensitive to variations in fuel flow at peak EGT than it is to variations in fuel flow at the test point. If the recorded fuel flow at peak EGT is too high, because the fuel flow system was reading high, that causes the calculated power at mixture for best power to be high. But, it also gives the impression that you are further LOP than you really are, so that means there is a bigger decrement from this too high calculated power at mixture for best power. In other words there are two effects that cancel each other out by about 80%. It is important to get a fairly accurate power at the steady state test point, but there we have the luxury of being able to spend a few minutes to get a good average reading.

Another potential issue is how quickly EGT indications respond when adjusting the mixture to find the fuel flow for peak EGT. My EGT indications are fairly sluggish - perhaps my EGT probes are further from the exhaust port than optimum. So I need to really take my time when looking for the fuel flow for peak EGT.

My third big concern is the effect of electronic ignition. The data that was used to produce this test method would have been from engines with two mags. It may tend to calculate a too low power for engines with EI at lower power settings. If I get enough data from aircraft with EI, perhaps there may be a trend that will suggest an empirical increment to the iSFC curves at low power conditions.

I'm looking forward to seeing what the data tells us.

Fuel Flow Variation: Not just the carb float..
 

Kevin, mine is muti-port FI and it does the same thing. I probably have the same sensor that Paul has since we both have GRT's. I have the best location (on the floor forward of the FI pump, aft of the firewall) that I can from the point of view of straight runs coming and going, but in that location the mechanical pump is sucking gas through it, not pushing it through. Anyhow, I think that the carb float is a good explanation, but not completely sufficient.

For those with EFIS's, it may be possible to use the gallons-used field that GRT and likely others provide. That and a stopwatch might solve the accuracy problem. I've not tried it yet, but I might, someday. I'm disqualified from your study with my FP prop.

I just took a look at my "decode" XLS from recording my GRT readings in flight. It has a very precise flight timer and a fuel flow reading. A little work with Excel (or OpenOffice) can improve your data considerably if your testers use GRT or equivalent. Averaging looks feasible.

Earlier in this thread I said that I only wanted data from aircraft with CS props, as the method is designed around a constant rpm. But, I have an idea of a possible way to extend this method for FP props. I'll have to crank out a new test card, and a slight change to the test procedure to gather rpm data at each fuel flow. Stand by a day or two and I'll open the doors to FP planes.

Have data to send you.
 

Kevin - Check PM.