• September 25, 2018, 11:48:14 AM
• Welcome, Guest

Pages: 1 [2] 3   Go Down

### AuthorTopic: Modeling fuel consumption using Lockheed 487 Report.  (Read 33551 times)

#### Richard C Cooke

• T1
• Posts: 35
##### Re: Modeling fuel consumption using Lockheed 487 Report.
« Reply #15 on: April 01, 2012, 02:27:01 PM »

Contrary to Lockheed report 487 and other documents that state the BSFC of .42, the best BSFC obtained according to this Pratt and Whitney table (the people who manufactured the engines) is .48 and this was at 300 and 350 hp. Cruising with 350 hp per engine burns 28 gallons per hour per engine which would have used 1100 gallons in 19:38. Cruising with 300 hp per engine uses the 1100 gallons in 22:56.

The lowest fuel flow stated is 21 gallons per hour per engine for 250 BHP (BSFC of .50) so it appears that Earhart's statement on the Hawaii flight that they were burning 20 gallons per hour was referring to the fuel flow to each engine making the total fuel flow about 40 gallons per hour. At 21 gallons per hour per engine the 1100 gallons would have produced an endurance 26:11. (But higher power settings had to be used early in the flight.)
gl
The P&W chart fuel consumption is a lot higher than Kelly Johnson's tests results for Earhart's plane.

His lowest test consumption was 18 gallons per hour at 1700rpm and 22.0 in manifold pressure.  To estimate the test bhp I went from the P&W chart that, at 1700rpm, gives 250bhp at 19.0 in HG and 350bhp at 27.0 in HG.  So 22.0 in HG should produce @ 285-290bhp, which would give an incredible BSFC of 0.38.

The only suggestion for the difference I have is that the P&W spec sheet is very conservative on fuel consumption, maybe because they did not want customers to run out of fuel.  There are examples of engines being run on weak mixtures in WWII so it was clearly possible to beat the standard book values.  The planes for the Doolittle raid had their engines set to very lean mixtures to extend their range and several were apparently reset to standard settings by helpful Hornet mechanics which unfortunately caused them to use up their fuel too quickly.

rc
Logged

#### Gary LaPook

• T5
• Posts: 1624
##### Re: Modeling fuel consumption using Lockheed 487 Report.
« Reply #16 on: April 03, 2012, 12:25:59 AM »

The P&W chart fuel consumption is a lot higher than Kelly Johnson's tests results for Earhart's plane.

His lowest test consumption was 18 gallons per hour at 1700rpm and 22.0 in manifold pressure.  To estimate the test bhp I went from the P&W chart that, at 1700rpm, gives 250bhp at 19.0 in HG and 350bhp at 27.0 in HG.  So 22.0 in HG should produce @ 285-290bhp, which would give an incredible BSFC of 0.38.

The only suggestion for the difference I have is that the P&W spec sheet is very conservative on fuel consumption, maybe because they did not want customers to run out of fuel.  There are examples of engines being run on weak mixtures in WWII so it was clearly possible to beat the standard book values.  The planes for the Doolittle raid had their engines set to very lean mixtures to extend their range and several were apparently reset to standard settings by helpful Hornet mechanics which unfortunately caused them to use up their fuel too quickly.

rc

So, there is a conflict in the data. If the P&W power output numbers are right then the BSFC, based on Johnson's fuel flow data, is unbelievable low.  If the Johnson data is correct then the power output data from P&W must be wrong.

Who to believe, who to believe?

I'm going with P&W, the people who made the engines. It takes years to develop aircraft engines with a gazillion hours of testing. Kelly Johnson (one of the greatest aircraft designers of all time) had a much more limited opportunity to test the engines so had a much more limited set of data to work with. I represented P&W for many years and many of the cases involved the same engine as Earhart's, which were commonly used in crop dusters, plus a lot of PT-6 cases. In the '30s these engines were used on the airliners of the day, such as the Electra, and airlines are very conscious of operating costs. If P&W could have touted that their engines could get a better BSFC, which results directly in lower aircraft operating costs, they could have sold more engines so they had a strong business purpose to state the lowest BSFC that they could get away with and they best they claimed was 0.46 which is also the lowest claimed on page 33 of Lockheed report 487, attached.

Several years ago I worked on a case involving a Piper Saratoga SP II which uses a Lycoming TIO-540-AH1A turbocharged engine of 300 h.p. and I studied the very complete and voluminous engine testing documents produced by Lycoming. At cruise power this engine has a BSFC of .45 while at full power the BSFC increases to .65, the same as Earhart's.  Although it is not a round engine like the S3H1 it is a much newer design, being certificated in 1997, so it is unlikely that Earhart's 1930's engine got better fuel consumption. You can  make the mixture just so lean before the fire goes out so there is a limit to how far you can stretch the plane's endurance.

gl

« Last Edit: April 03, 2012, 10:36:43 AM by Gary LaPook »
Logged

#### Richard C Cooke

• T1
• Posts: 35
##### Re: Modeling fuel consumption using Lockheed 487 Report.
« Reply #17 on: April 03, 2012, 03:56:11 PM »

The discrepancy comes from what bhp you get from a combination of rpm and manifold pressure and altitude.  I had not adjusted the bhp for altitude hence the incredible BSFC, sorry about that.

The key point is that Kelly Johnson measured 18 gallons per hour per engine, so any minimum fuel consumption would be that or less.

rc
« Last Edit: April 07, 2012, 01:44:09 PM by Richard C Cooke »
Logged

#### JNev

• T5
• Posts: 778
• It's a GOOD thing to be in the cornfield...
##### Re: Modeling fuel consumption using Lockheed 487 Report.
« Reply #18 on: April 05, 2012, 04:43:26 AM »

I have 'no problem' with it but am not certain what it gains, one way or the other.

I do believe Johnson's data was thorough and correct, and that AE should have been able to use the analyzer, etc. effectively enough to achieve good results.  Johnson had flown with her, for one thing, and despite many of her shortcomings I don't think was a complete fool.  It's not that hard to exercise reasonable fuel management.

For my understanding, the Johnson data - especially considering the 3 telegrams - is ample to understand the fuel consumption characteristics of the airplane and what guidance AE had to work with.

LTM -
- Jeff Neville

Former Member 3074R

Logged

#### Gary LaPook

• T5
• Posts: 1624
##### Re: Modeling fuel consumption using Lockheed 487 Report.
« Reply #19 on: April 08, 2012, 01:43:00 AM »

I'll stick to Kelly Johnson's figures, personally.

The three memos are reprinted therein.  I suggest the reading as they lend more detail and background than just looking at the Lockheed report can provide.

Draw your own conclusions, of course, but I personally respect what was derived via testing as was done, and overseen by Pratt & Whitney, as was done.

LTM -
We may be talking past each other. The Kelly Johnson telegrams do not provide enough information to allow us to calculate the BSFC of the engine since, although they provide the rate of fuel flow, they do not state a power output for the fuel flows. Nor do they tell us the airspeed achieved so we can't even take a shot a computing the power outputs. So if, for example, the 36 gph fuel flow was only putting out 468 hp then the BSFC was the expected and normal 0.46. (It is easy to use the rule of thumb that one gallon per hour buys you thirteen horsepower.)
I still doubt that a round carburetted  1930s engine with a geared supercharger could get better fuel specific consumption than a 60 year newer design, the 1997 Lycoming TIO-540-AH1A engine made with modern materials, has fuel injection and a non-power robbing turbo-charger and it only gets 0.45.

gl
« Last Edit: April 08, 2012, 01:59:19 AM by Gary LaPook »
Logged

#### Gary LaPook

• T5
• Posts: 1624
##### Re: Modeling fuel consumption using Lockheed 487 Report.
« Reply #20 on: April 11, 2012, 02:08:42 AM »

The point of the telegrams is to add practical detail derived from experience - it seems to have been a fine-tuning of the data effort by Johnson himself.  Specifics may be derived by going to the report, etc. to correlate the numbers called out and extrapolate the data in detail.  In fact, the telegrams alone do go a long way toward giving workable calculations as to what techniques could be practically applied and what the results should be for a trip such as Lae - Howland, at least in a gross sense.  You don't really have to go too far into the weeds to see what endurance the Electra had given the observations of Johnson.

Remember, Report 487 was prepared specifically for the flight from Hawaii non-stop to Tokyo so the maximum range was very important because it is 3850 SM flight. The range with the normally expected BSFC was 4100 SM giving only a 250 SM reserve. With the optimistic BSFC the claimed range was 4500 SM which would have provided a much more reasonable 650 SM reserve, 17% reserve. Even Earhart wasn't buying it and obviously didn't believe this claimed range since she tried to set up an aerial refueling over Midway and that plan went forward until the Howland option was created. For the flight from Lae to Howland range was not critical, she did not run out of fuel on the way to Howland, she ran out of gas a minimum of one hour after arrival in that vicinity and possibly three of four hours later. There was no reason for her to fly the plane in a way to get the maximum possible range and we know that she didn't since she flew at a much higher airspeed than optimal for range. She averaged 133 mph ground speed into about a 27 mph headwind so she was flying at an airspeed of about 160 mph which was faster than optimal even at the maximum weight of 16,500 pounds and way above the optimum 120 mph at the final weight. Review pages 30 and 34 of the report and you will see that her speed was too high. Page 34 shows the best range speed at 16,500 lbs is 144 mph and 135 at 9,300 lbs. Doing some computations on page 30, (mph/hp required,) shows 150 mph at maximum weight and 120 mph at 9,300 lbs.

The Kelly Johnson telegrams addressed the much shorter flight to Hawaii so the numbers in them do not relate to the 487 scenario. Since the telegrams do not tell us anything about the power output or airspeed there is no reason to believe these numbers produce a better BSFC than the standard 0.46. What it appears the telegrams are telling Earhart is that if you use these power settings then this is the fuel flow that you will get. Then it is up to you to see what airspeed they produce and determine your specific range and it you have a comfortable safety margin. Nowhere do the telegrams state "that if you use these settings you will achieve a 0.42 BSFC" because such fuel economy was no longer needed for the much shorter legs in the revised flight plan.
Quote

I can understand why it may be hard to believe that a 'modern' flat engine is no more efficient than an 'old' round engine - but we're still talking pretty old technology in terms of the engine itself: air cooled clunky compared to truly modern engines (an affliction that plagues 'modern' light aircraft to this day compared to the advances made in recent decades by the automotive industry).  The fuel injection helps, but turbos don't run for free, either.  And, what 'modern' materials and what differences do they make?  I guess we commonly have ceramic cylinder coatings, etc. in some applications, for sure - but how much have those things really improved specific fuel consumption?  As to the magic of fuel injection - factory stuff is notoriously poorly balanced for flow and consumption numbers are generally conservative as a result in my experience.

And has anyone that you know of really rung-out a TIO-540-AH1A to get 'real' figures in a given installation the way that Johnson & Company did for the Electra?  Maybe the results WOULD be surprisingly efficient, especially if they were working to optimize the procedures to get consistent results, etc.  That may be one big difference - I'm convinced that Johnson was not day-dreaming; maybe an equal amount of experimentation with the Lycoming would yield something similar in terms of economoy.

On my first trip to Florida I bought avgas for 16 cents a gallon now it's \$5.99 a gallon. I don't remember what it cost in 1997 when the TIO-540-AH1A was certificated but it was a lot more than 16 cents a gallon. Fuel  is the major direct operating cost and a better BSFC would have been an important marketing advantage for Lycoming so you can be sure they used all the new technology available sixty years after Earhart to make that engine fuel efficient and they could only get it to 0.45.
gl
« Last Edit: April 11, 2012, 04:28:34 PM by Gary LaPook »
Logged

#### Richard C Cooke

• T1
• Posts: 35
##### Re: Modeling fuel consumption using Lockheed 487 Report.
« Reply #21 on: April 13, 2012, 09:56:16 AM »

It might be that the TIO-540-AH1A in not that good on specific fuel consumption.  Here is an economy cruise chart for a Mooney 201 with the best specific coming from 68% (136bhp) at 2200rpm giving a fuel consumption of 55.5lb/hr which is 0.408 lb/bhp-hr.

RC
Logged

#### JNev

• T5
• Posts: 778
• It's a GOOD thing to be in the cornfield...
##### Re: Modeling fuel consumption using Lockheed 487 Report.
« Reply #22 on: April 13, 2012, 09:05:38 PM »

I personally can't give high marks to 'modern' air cooled engines for peak economy and don't see a remarkable difference in one manufactured today vs. 50 years ago, or even 75 years ago: you can only realize so much coefficient anything out of all that slop.  Tolerances are about the same by their nature - about twice the clearance throughout the reciprocating parts when compared to their liquid-cooled cousins.

You've done a better job Richard at finding and posting some facts on that than I am putting up here, of course, but I am agreeing with your point.

On the other hand, my F150 pickup has a 4.8L V-8 that gives astonishing torque and mileage - and is holding true for 130K miles and counting.  It delivers far better performance than my old trusty F100 300 CID straight six did (4.9L) at noticeably better economy.

I just can't sense anything of the sort in the 'evolution' of air cooled aircraft engines - we're flying very old technology there, however reliable.

Good find.

LTM -
- Jeff Neville

Former Member 3074R

Logged

#### Richard C Cooke

• T1
• Posts: 35
##### Re: Modeling fuel consumption using Lockheed 487 Report.
« Reply #23 on: April 14, 2012, 05:00:56 PM »

My chart came from a book I bought 25 years ago when I was thinking of buying a Mooney.  I also have a book on Pipers, where the best sfc I can find is for the PA38 Tomahawk, which they claim consumes 4.9 gph at 65% of its 112bhp.  That works out as 29.4 lb/hr for 72.8 bhp, giving a BSFC of 0.404 lb/bhp-hr.

Going back to the Electra I noticed that one of Kelly Johnson's tests is quite close to one of the points on the chart on page 33/34 of L487.  His test gives 43 gph at 5000ft, 1800rpm and 26.0 in HG, and L487 gives 300 bhp at 5000ft, 1800rpm and 25.8 in HG.  So if we ignore the difference between 25.8 and 26.0 in HG, we have a 300bhp setting that uses 43 gph for both engines, and works out to a BSFC of 0.43 lb/bhp-hr.

RC
Logged

#### Richard C Cooke

• T1
• Posts: 35
##### Re: Modeling fuel consumption using Lockheed 487 Report.
« Reply #24 on: April 15, 2012, 09:03:35 PM »

Re reading the words in L487 it says that the 0.46 BSFC is "conservative" and the 0.42 BSFC is what you need to achieve to get a 4500 mile range.  Note that they don't say they tested the engines and got that, just that this is what the engines need to achieve to get a 4500 mile range.  The Kelly Johnson test achieving 0.43 suggest they were only slightly optimistic.

RC
Logged

#### William Thaxton

• T1
• Posts: 20
##### Re: Modeling fuel consumption using Lockheed 487 Report.
« Reply #25 on: April 22, 2012, 08:07:16 PM »

I guess I'm the new guy on the block but here are a couple of points from my experience which might be of interest to this discussion.  First, my experience:  a bit over 8000 hours of transport time with about 1800 of those hours being in a T-29/Convair 440 powered by twin R2800-99W radial engines.

First point:  we used an unpublished (but generally accepted and used) leaning procedure to extend our range during cruise flight.  This consisted of setting mixture control to autolean, setting cruise power for the desired airspeed, carefully syncing the props, then leaning the mixture below autolean slowly until we got a 100 RPM drop on each engine.  This procedure extended our range/endurance by about 10%.  As fuel burned off and power was reduced to maintain a constant airspeed we would repeat the procedure at least once an hour.

Second, I haven't seen much discussion of temperature in this thread and temperature can be critical to the operation of a normally aspirated engine, especially in the low altitude environment (more extreme variation than in the high altitude environment).  Hot air is less dense and requires a higher throttle setting to achieve a similar effective power delivered at the prop.  I still remember the loss of an engine over the Mojave Desert one summer day while operating at about 1500 AGL (ambient temp about 120 degrees).  My remaining engine simply couldn't develop enough power to maintain altitude but I was able to hit the overrun at Edwards AFB about 500' short of the threshold.  That definately made the effect of temperature a priority subject in my mind.  I was reminded of this by the report of clearing brush on Gardner (Nikumaroro) in 120 degree temps as reported in Niku VI.  Recips just don't like hot air!

Just some random thoughts which I hope will be of help.

William
« Last Edit: April 23, 2012, 07:50:17 AM by William Thaxton »
Logged

#### Gary LaPook

• T5
• Posts: 1624
##### Re: Modeling fuel consumption using Lockheed 487 Report.
« Reply #26 on: April 23, 2012, 01:02:44 AM »

Second, I haven't seen much discussion of temperature in this thread and temperature can be critical to the operation of a normally aspirated engine, especially in the low altitude environment (more extreme variation than in the high altitude environment).  Hot air is less dense and requires a higher throttle setting to achieve a similar effective power delivered at the prop.  ...I was reminded of this by the report of clearing brush on Howland in 120 degree temps.  Recips just don't like hot air!

Just some random thoughts which I hope will be of help.

William
But the air temperature over the sea is very constant. In the vicinity of Gardner in July the temps are between 76° and 79° F 10% of the time, between 79 and 84 85% of the time and between 84 and 86 the remaining 5% of the time. It is never above 86° F.

If you are thinking about her takeoff performance back at Lae you should look at this prior post.

You mentioned "naturally aspirated engines" but Earhart's S3H1s were supercharged as was your R2800-99W. BTW, did you try your water injection on your narrow scrape at Edwards?

gl
« Last Edit: April 23, 2012, 01:16:38 AM by Gary LaPook »
Logged

#### William Thaxton

• T1
• Posts: 20
##### Re: Modeling fuel consumption using Lockheed 487 Report.
« Reply #27 on: April 23, 2012, 08:29:37 AM »

Marty, thanks for the "modify" info.  Like I said "New Guy".

Gary,
Not trying to be argumentative but the temps you list (while intuitively accurate based on my knowledge of other Pacific atols and islands [Diego, Guam, etc]) don't seem to jive with the Niku VI report.  In any case, just another thought to be used as food for the discussion.

As for use of ADI, that was my "ace" at Edwards.  Our normal procedure was to use water for the initial takeoff, then shut it off quickly to conserve water for just such an instance.  Total ADI capacity was about 5 minutes at max power so that normally we had about a 2 1/2 to 3 minute "reserve" available.  While ADI does boost the engine horsepower (about 200 HP per engine for the R-2800) it also put a significant strain on the operating parts and sharply reduces the engine life curve.  Since I was already down to my last operating engine I elected not to use ADI once I saw I could hit the overrun.  That was also one of the factors in selecting Edwards.  A tap on the rudders and I could use Murock Dry Lake.  Good enough for the X-15 was good enough for me!  Also, you are correct about the "supercharged" engines but the superchargers weren't normally used in the low altitude environment.  In fact, procedurally, we never engaged the blowers below 14,000'.  Again, it was that pesky engine life curve.  The superchargers helped "trick" the engine into thinking it was at a lower altitude by packing air into the carburetor.  When used at low altitude this could easily blow a cylinder.  While I suppose there could be an emergency situation where such use might be approporiate it certainly would not be a normal course of events.  This would indicate that, in normal flight at low altitude, AE would have been flying a normally aspirated aircraft and the supercharger would not have been engaged.

In any case, my original thought was that the cruise leaning procedure we used would bring the range/fuel consumption estimates much closer to those presented by Kelly Johnson and might provide an explanation for the apparent differences.  The temperature question, and that is what it really is, was more of an afterthought.

Greetings to all,
William
Logged

#### Martin X. Moleski, SJ

• Posts: 2910
##### Re: Modeling fuel consumption using Lockheed 487 Report.
« Reply #28 on: April 23, 2012, 08:46:33 AM »

Marty, thanks for the "modify" info.  Like I said "New Guy".

No problem.  Glad to have you with us.

Quote
... Also, you are correct about the "supercharged" engines but the superchargers weren't normally used in the low altitude environment.  In fact, procedurally, we never engaged the blowers below 14,000'.  Again, it was that pesky engine life curve. ...

I've never heard anyone discuss AE's protocol for use of the superchargers.

It seems to me that she had an effective service ceiling of 10,000 feet--at least, if I remember correctly, that was as high as she intended to go in her flight planning.  No oxygen system on the aircraft.

If her superchargers were intended only for above 14,000 feet, then it seems that they were pretty much useless for her purposes.
LTM,

Marty
TIGHAR #2359A

Logged

#### Richard C Cooke

• T1
• Posts: 35
##### Re: Modeling fuel consumption using Lockheed 487 Report.
« Reply #29 on: April 23, 2012, 11:13:55 AM »

If her superchargers were intended only for above 14,000 feet, then it seems that they were pretty much useless for her purposes.
She needed the superchargers at take off to get the 36 in HG manifold pressure needed for 600bhp.

RC
Logged
Pages: 1 [2] 3   Go Up

 Copyright 2018 by TIGHAR, a non-profit foundation. No portion of the TIGHAR Website may be reproduced by xerographic, photographic, digital or any other means for any purpose. No portion of the TIGHAR Website may be stored in a retrieval system, copied, transmitted or transferred in any form or by any means, whether electronic, mechanical, digital, photographic, magnetic or otherwise, for any purpose without the express, written permission of TIGHAR. All rights reserved. Contact us at: info@tighar.org • Phone: 610-467-1937 • Membership form • webmaster@tighar.org