X-Virus-Scanned: clean according to Sophos on Logan.com Return-Path: Sender: To: lml@lancaironline.net Date: Sat, 31 Jan 2009 07:39:25 -0500 Message-ID: X-Original-Return-Path: Received: from QMTA09.westchester.pa.mail.comcast.net ([76.96.62.96] verified) by logan.com (CommuniGate Pro SMTP 5.2.12) with ESMTP id 3460435 for lml@lancaironline.net; Fri, 30 Jan 2009 10:17:53 -0500 Received-SPF: none receiver=logan.com; client-ip=76.96.62.96; envelope-from=gregw@onestopdesign.biz Received: from OMTA09.westchester.pa.mail.comcast.net ([76.96.62.20]) by QMTA09.westchester.pa.mail.comcast.net with comcast id 9n9V1b0050SCNGk59rHJaM; Fri, 30 Jan 2009 15:17:18 +0000 Received: from osd1 ([24.6.40.29]) by OMTA09.westchester.pa.mail.comcast.net with comcast id 9rHG1b00g0dkeQQ3VrHHSQ; Fri, 30 Jan 2009 15:17:18 +0000 X-Original-Message-ID: <5137FCA375C84925BD416AB383772E8C@yosemite.onestopdesign.biz> From: "Greg Ward" X-Original-To: "Lancair Mailing List" References: Subject: Re: [LML] Re: Alternative Power Plants X-Original-Date: Fri, 30 Jan 2009 07:19:15 -0800 MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_0029_01C982AB.0EDB4780" X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2900.5512 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.5579 This is a multi-part message in MIME format. ------=_NextPart_000_0029_01C982AB.0EDB4780 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi Brent; Thanks for your comprehensive input below. You do bring up a couple of = issues that we have only looked at in passing, and we will look into = these issues further. Our engine builder/fabricator is Bob Wirth Racing = Engines of Hayward, CA, who has extensive experience with the 12A and = 13B over the years, and one of the reasons we have taken so long in = developing what you saw in the photos, is due to the research you speak = of. I'll try to address the points below individually: Greg ----- Original Message -----=20 From: Brent Regan=20 To: lml@lancaironline.net=20 Sent: Thursday, January 29, 2009 8:23 PM Subject: [LML] Re: Alternative Power Plants Hello Greg, Thank you for posting the pictures of the 20B aircraft installation = you are fabricating. Your pictures and my personal past experience = compels me to offer some words of caution. First, let me say that I = have no idea as the the degree and quality of the engineering that has = gone into this project to date. I am operating on a worst case = assumption so please discard comments concerning issues that you have = already addressed. I hope to present these in a self evident fashion. = Know that I have been involved in rotary engine development since the = late '70s, that I was involved in developing and building a 12A engine = that competed and won it's class in the 24 hours of Daytona endurance = race, that I have previously developed high output rotary engines for = aerospace applications, that I worked with Charlie Jones at Curtis = Wright to help market their SCORE aircraft rotary engine to Cessna and = Piper and that I seriously evaluated the Mazda R26B engine (4 rotor = winning Le Mans in '91, see SAE paper #920309) for my Lancair IV-P but = opted to adapt a Lycoming TIO540. I flew that airplane, and won, the = Denver to Oshkosh race in 1996. I now have over 1,000 hours on the = original engine, including cylinders, pistons, rings and valves.=20 The Wankel rotary engine we know is actually a compromise from Felix = Wankel's original concept. The progenitor engine, designated DKM-54, = featured a rotating rotor and housing arraigned such that each had a = pure rotary motion. It was NSU's engineer Walter Froede that came up = with the internally geared epicyclic design we are familiar with today. = While this design is attractive due to its high power density and = minimal parts count, it is not without its problems. One of the = greatest problems is the high aspect ratio of the combustion volume. = This leads to high percentage of charge quench and therefore poor brake = specific fuel economy, high hydrocarbon emissions and high exhaust gas = temperature. The CW SCORE engines attempted to overcome these = shortcomings with a heavy fuel direct injection with spark assist. They = were moderately successful. Another problem is the high power density with respect to heat = rejection. Nominally 20 percent of the total combustion energy must be = rejected to the cooling system. For a 300 Hp engine this heat flux is = more than 100 kW, enough to run 20+ average single family homes. This is = divided between the oil and coolant so the oil is at risk of chemical = breakdown above 240F (natural oils). I solved this problem on the race = engines by adding a tube shell intercooler between the oil and water = cooling systems that limited the oil temperature to 230F by allowing the = coolant to boil in the heat exchanger, dramatically raising the heat = transfer rates. In short, your three biggest problems will be heat, heat = and heat. Comments referencing your photos and subsequent post: 1) Belly mounted cooling system. The engine is VERY sensitive to air = in the coolant. You MUST have a de-aeration system mounted above the = highest point of the engine. This is a convenient point to place a = "radiator cap". ***We have an expansion tank mounted high on the firewall to handle = this very issue. There has been much conversation on the Rotary list = about "burping" the engine to make sure there is no air in the system. 2) Oil sump. It appears that you are using a stock oil sump = configuration. You will find this insufficient for the anticipated = continuos power levels. You need a high volume (12 Qt) remote tank dry = sump system. The oil tank should have a centrifugal de-aeration design = with multiple baffles. You should use a synthetic oil that is more = tolerant to high temperature operation. ***Actually, we are using a dual system. We are using the stock sump = for the engine, and an aux. small tank for turbo and redrive oiling, = with a pump designed to return this oil back to the system (modified = dry-sump). We have a large oil cooler in the belly scoop, cooled = separately from the radiator. This system is identical to the P-51 = scoop configuration. We are using AN-10 tubing/hose all the way, with = dual large oil filters. With all of this plumbing and cooling capacity, = we figure we have about doubled the oil capacity over stock, dyno and = temps will tell. 3) PSRU. It appears that you are using a planetary gear reduction, = likely adapted from an automotive automatic transmission. Be advised = that this configuration will generate between 1% and 2% of the = transmitted power in heat, or about 5Kw at 300 Hp. You need a separate = oil circulation and cooling system for this reduction. Prior attempts to = use engine oil have been marginally successful, at best. I am also = assuming that you have sufficient torsion compliance between the engine = and gears to prevent cyclical load fatigue failure. ***The redrive is Tracy Crook's design, which has been well tested = both on 13B's and 20B's. It is rated at 450 hp at take-off, then we = pull back to 55-70% at cruise. As I said above, it has a separate = oiling system from the engine. 4) Induction Plenum. It appears you are planning single port injection = based on mass flow into the intake plenum. It is important that the flow = distribution of the plenum be tested, otherwise you may have poor charge = and mixture uniformity across the three rotors. Flow testing can be = accomplished with a shop vac and carburetor flow gage (balance). ***Actually, each rotor is configured with 2 injectors. One is in the = stock location on the block, the other is in the manifold we have built = (see picture). Using Tracy's EC-3, we will map this on the dyno and = make the necessary adjustments. Bob Wirth is one of the best flow bench = guys in the State, and with much consultation with Tracy and others, we = feel we have gotten close to the optimum design, dyno will tell. 5) Trochoid Cooling. Stock rotor housings have poor cooling = distribution for high continuos power outputs. You need to have = modified rotor housings and a high output water pump that is driven at = the optimal speed. ***Not sure on this one. We will dyno with the stock pump, and see = what happens. We have even talked about reducing the pump speed, for = cavitation issues. 6) High EGT. While having individual exhaust manifolds to the turbo, = to utilize the kinetic energy of the exhaust, is good, these should be = fabricated from Inconel to survive the high EGTs without failure. = Likewise, the turbo needs to be a high temperature version. ***We did schedule 340, because of the tendency of Iconel to need to = be at constant hight temps to perform correctly. With our envelope, 340 = seemed the best and cost effective approach. Again, dyno time and temps = will tell. 7) Engine Mount. It is not apparent why the engine mounts are = vertical. Thrust loads from the propeller can approach 1,000 lbs. The = mounts, as designed, will not respond well to these loads. ***We are in the process of beefing the mount now. We weren't happy = with the way this one came out. The pan plate and front plate were = designed to work together, and have been well tested. No failures to = date.=20 8) PSRU mount. While I can appreciate that a heavy flat plate is easy = to fabricate, it is poorly suited to react the gyroscopic precession = loads and moments generated by the propeller. You would be wise to = perform an independent engineering analysis of this design. ***Again, this unit has several years and a lot of installations with = no failures. We will keep a close eye on it. 9) Transverse Engine Mount Triangulation. It is not clear from the = photos but it seems like there is insufficient bracing to react = transverse nose gear loads. Ask Skip Slater about this one. ***Not sure on this one. Will do some checking. It is the stock = Lancair Continental bed mount, modified to accept the Rotary, and then = we beefed the hell out of it, and intend on adding some additional = bracing to it. We will send pictures of the final mess when we get it = mounted back on the firewall....... 10) Reversed slip couplings. The exhaust slip couplings seem to be = reversed. Upstream should be the interior tube. ***This is the way Burns configured it. We will weld tabs and bolts = to hold them together.=20 The above list is by no means comprehensive. It is simply a free = association exercise based on the posted photos. Engine installation = design is serious business. There is a long list of designers that have = died behind their engines. Success is in no small part dependent on the = ability to understand and allow for the errors of those who have gone = before. Remember that almost any engine can power an aircraft. The = quality of the engineering and implementation of the installation = determines the longevity of the application and, sadly, frequently the = pilot. Fly safe. Brent Regan ------=_NextPart_000_0029_01C982AB.0EDB4780 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hi Brent;
Thanks for your comprehensive input = below.  You=20 do bring up a couple of issues that we have only looked at in passing, = and we=20 will look into these issues further.  Our engine builder/fabricator = is Bob=20 Wirth Racing Engines of Hayward, CA, who has extensive experience with = the 12A=20 and 13B over the years, and one of the reasons we have taken so long in=20 developing what you saw in the photos, is due to the research you speak=20 of.  I'll try to address the points below=20 individually:
Greg
----- Original Message -----
From:=20 Brent=20 Regan
Sent: Thursday, January 29, = 2009 8:23=20 PM
Subject: [LML] Re: Alternative = Power=20 Plants

Hello Greg,

Thank you for posting the = pictures of=20 the 20B aircraft installation you are fabricating. Your pictures and = my=20 personal past experience compels me to offer some words of = caution. =20 First, let me say that I have no idea as the the degree and quality of = the=20 engineering that has gone into this project to date. I am operating on = a worst=20 case assumption so please discard comments concerning issues that you = have=20 already addressed. I hope to present these in a self evident fashion. = Know=20 that I have been involved in rotary engine development since the late = '70s,=20 that I was involved in developing and building a 12A engine that = competed and=20 won it's class in the 24 hours of Daytona endurance race,  that I = have=20 previously developed high output rotary engines for aerospace = applications,=20 that I worked with Charlie Jones at Curtis Wright to help market their = SCORE=20 aircraft rotary engine to Cessna and Piper and that I seriously = evaluated the=20 Mazda R26B engine (4 rotor winning Le Mans in '91, see SAE paper = #920309) for=20 my Lancair IV-P but opted to adapt a Lycoming TIO540.  I flew = that=20 airplane, and won, the Denver to Oshkosh race in 1996.  I now = have over=20 1,000 hours on the original engine, including cylinders, pistons, = rings and=20 valves.

The Wankel rotary engine we know is actually a = compromise from=20 Felix Wankel's original concept. The progenitor engine, designated = DKM-54,=20 featured a rotating rotor and housing arraigned such that each had a = pure=20 rotary  motion. It was NSU's engineer Walter Froede that came up = with the=20 internally geared epicyclic design we are familiar with today.  = While=20 this design is attractive due to its high power density and  = minimal=20 parts count, it is not without its problems.  One of the greatest = problems is the high aspect ratio of the combustion volume. This leads = to high=20 percentage of charge quench=20 and therefore poor brake specific fuel economy, high hydrocarbon = emissions and=20 high exhaust gas temperature.  The CW SCORE engines attempted to = overcome=20 these shortcomings with a heavy fuel direct injection with spark = assist. They=20 were moderately successful.

Another problem is the high power = density=20 with respect to heat rejection. Nominally 20 percent of the total = combustion=20 energy must be rejected to the cooling system. For a 300 Hp engine = this=20 heat  flux is more than 100 kW, enough to run 20+ average single = family=20 homes. This is divided between the oil and coolant so the oil is at = risk of=20 chemical breakdown above 240F (natural oils). I solved this problem on = the=20 race engines by adding a tube shell intercooler between the oil and = water=20 cooling systems that limited the oil temperature to 230F by allowing = the=20 coolant to boil in the heat exchanger, dramatically raising the heat = transfer=20 rates. In short, your three biggest problems will be heat, heat and=20 heat.

Comments referencing your photos and subsequent = post:

1)=20 Belly mounted cooling system. The engine is VERY sensitive to air in = the=20 coolant. You MUST have a de-aeration system mounted above the highest = point of=20 the engine. This is a convenient point to place a "radiator=20 cap".
***We = have an expansion=20 tank mounted high on the firewall to handle this very issue.  = There has=20 been much conversation on the Rotary list about "burping" the engine = to make=20 sure there is no air in the system.


2) Oil sump. It appears that you are using a stock oil = sump=20 configuration. You will find this insufficient for the anticipated = continuos=20 power levels. You need a high volume (12 Qt) remote tank  dry = sump=20 system. The oil tank should have a centrifugal de-aeration design with = multiple baffles. You should use a synthetic oil that is more tolerant = to high=20 temperature operation.
 
***Actually, we = are using a=20 dual system.  We are using the stock sump for the engine, and an = aux.=20 small tank for turbo and redrive oiling, with a pump designed to = return this=20 oil back to the system (modified dry-sump).  We have a large oil = cooler=20 in the belly scoop, cooled separately from the radiator.  This = system is=20 identical to the P-51 scoop configuration.  We are using AN-10=20 tubing/hose all the way, with dual large oil filters.  With all = of this=20 plumbing and cooling capacity, we figure we have about doubled the oil = capacity over stock, dyno and temps will = tell.

3) PSRU.=20 It appears that you are using a planetary gear reduction, likely = adapted from=20 an automotive automatic transmission. Be advised that this = configuration will=20 generate between 1% and 2% of the transmitted power in heat, or about = 5Kw at=20 300 Hp. You need a separate oil circulation and cooling system for = this=20 reduction. Prior attempts to use engine oil have been marginally = successful,=20 at best. I am also assuming that you have sufficient torsion = compliance=20 between the engine and gears to prevent cyclical load fatigue = failure.
 
***The redrive is = Tracy=20 Crook's design, which has been well tested both on 13B's and = 20B's.  It=20 is rated at 450 hp at take-off, then we pull back to 55-70% at = cruise. =20 As I said above, it has a separate oiling system from the=20 engine.

4) Induction Plenum. It appears you are = planning single port injection based on mass flow into the intake = plenum. It=20 is important that the flow distribution of the plenum be tested, = otherwise you=20 may have poor charge and mixture uniformity across the three rotors. = Flow=20 testing can be accomplished with a shop vac and carburetor flow gage=20 (balance).
 
***Actually, each = rotor is=20 configured with 2 injectors.  One is in the stock location on the = block,=20 the other is in the manifold we have built (see picture).  Using = Tracy's=20 EC-3, we will map this on the dyno and make the necessary = adjustments. =20 Bob Wirth is one of the best flow bench guys in the State, and with = much=20 consultation with Tracy and others, we feel we have gotten close to = the=20 optimum design, dyno will tell.

5)  = Trochoid=20 Cooling. Stock rotor housings  have poor cooling distribution for = high=20 continuos power outputs.  You need to have modified rotor = housings and a=20 high output  water pump that is driven at the optimal = speed.
 
***Not sure on = this=20 one.  We will dyno with the stock pump, and see what = happens.  We=20 have even talked about reducing the pump speed, for cavitation=20 issues.

6) High EGT. While having individual=20 exhaust  manifolds to the turbo, to utilize the kinetic energy of = the=20 exhaust, is good, these should be fabricated from Inconel to survive = the high=20 EGTs without failure. Likewise, the turbo needs to be a high = temperature=20 version.
 
***We did = schedule 340,=20 because of the tendency of Iconel to need to be at constant hight = temps to=20 perform correctly.  With our envelope, 340 seemed the best and = cost=20 effective approach.  Again, dyno time and temps will=20 tell.

7) Engine Mount. It is not apparent why = the=20 engine mounts are vertical. Thrust loads from the propeller can = approach 1,000=20 lbs. The mounts, as designed, will not respond well to these = loads.
 
***We are in the = process of=20 beefing the mount now.  We weren't happy with the way this one = came=20 out.  The pan plate and front plate were designed to work = together, and=20 have been well tested.  No failures to=20 date. 

8) PSRU mount. While I can = appreciate that=20 a heavy flat plate is easy to fabricate, it is poorly suited to react = the=20 gyroscopic precession loads and moments generated by the = propeller.  You=20 would be wise to perform an independent engineering analysis of this=20 design.
 
***Again, this = unit has=20 several years and a lot of installations with no failures.  We = will keep=20 a close eye on it.

9) Transverse Engine Mount=20 Triangulation.  It is not clear from the photos but it seems like = there=20 is insufficient  bracing to react transverse nose gear loads. Ask = Skip=20 Slater about this one.
 
***Not sure on = this=20 one.  Will do some checking.  It is the stock Lancair = Continental=20 bed mount, modified to accept the Rotary, and then we beefed the hell = out of=20 it, and intend on adding some additional bracing to it.  We = will=20 send pictures of the final mess when we get it mounted back on the=20 firewall.......

10) Reversed slip couplings. = The=20 exhaust  slip couplings seem to be reversed. Upstream should be = the=20 interior tube.
 
***This is the way Burns configured it.  We will weld tabs = and bolts=20 to hold them together. 

The above list is by no means=20 comprehensive. It is simply a free association exercise based on the = posted=20 photos. Engine installation design is = serious=20 business.  There is a long list of = designers that=20 have died behind their engines. Success is in no small part dependent = on the=20 ability to understand and allow for the errors of those who have gone = before.=20 Remember that almost any engine can power an aircraft. The quality of = the=20 engineering and implementation of the installation determines the = longevity of=20 the application and, sadly, frequently the pilot.

Fly=20 safe.

Brent=20 Regan




------=_NextPart_000_0029_01C982AB.0EDB4780--