For general interest, just in case the need arises.

 Cotronics makes a variety of high temperature Adhesives.

 Within the list of Ceramic Adhesives are Metallic adhesises as well.
 A general metal bond is Durabond 950 good to 1200 deg. F.
 The data sheet shows a comparison to epoxy.


 Cotronics, High temperature adhesives
 http://www.cotronics.com/vo/cotr/ca_1.htm
 http://www.cotronics.com/vo/cotr/ca_metallic.htm
 Rolf

 Thanks Rolf!

 Paul and I were talking about aluminum center and end housings
 this past wed. , the problem with aluminum housings is creating
 a durable surface for the rotor side seals to ride on. The oem
 cast iron housings are surface hardened and precision ground,
 creating a surface similar to the bore piston rings move within.
 The grinding process leaves minor surface scratches within which
 a thin film of oil is maintained.

 Aluminum is not tough enough to withstand the constant
 pressure of the rotor seals moving against it. Paul stated, and
 he will correct me if I heard him wrong, that Mazda used the
 process called Detonation Gun Coating on the16b prototype engine
 to resolve the wear problem. I believe he also stated that the
 Mazdatrix aluminum end and center housings were constructed
 similarly. I did'nt think to ask about what Mazda did with
 magnesium housings, which I believe were used on the 91 LeMans 4
 rotor.

 As my process of learning about the technical difficulties of
 creating rotary engines has progressed, I have considered as
 many ways as I could think of, to get what I want in my personal
 aircraft rotary engine. So this one issue has stayed with me;
 How do I create the lightweight housings using my own skills and
 equipment? I have no desire to spend the several thousand
 dollars required to get a product that might not work on
 aircraft.

 The ceramic adhesives may do the trick. I considered oven
 brazing cast iron plates to each side of a pre machined (for
 coolant and oil) aluminum center section. There are metals that
 can bond them together. The melting point of the metals is
 perhaps to low, at around 450f.

 With the ceramic adhesives, it appears a high strength bond
 between the iron plate, and the machined aluminum (or perhaps
 magnesium) center section may be possible. With an engine
 operating temperature approximating 190f, the bond strength of
 the ceramic adhesive exceeds 800psi.; and then gets higher as
 the temperature  increases.

 I am not certain, but it appears to me that any expansion
 contraction differences between the iron and aluminum, may not
 have a negative effect with a bond this strong; considering the
 bonding surfaces are wide enough. With coolant pressure never
 exceeding 50psi at 200f, seems like no problems could occur with
 the bond at the center of the housings where there is no crush
 pressure from the  housing through bolts.

 Has it been tried already?
 Can this resolve a significant problem of lightening the rotary engine?

 Ed Carver

 Sounds like it is worth a try Ed. If the water jacket is open
 on one side it can be machined on an NC mill. Use the adhesive
 to seal the various chambers to the wear plate. Here are some
 3D's,dwg and pictures on the subject. Richard Sohn uses several
 high temp O-rings.


 Paul Lamar


 Ed Carver,  I investigated using magnesium  in rotary end housings last
year. There are two magnesiun/aluminum/zirconium alloys that barely make
the
temperature requirements.  You need to remember that it is the metal
temperature not water temperature that is critical. Magnesium was wrong in
many ways particularly if fasteners need to be threaded into it where pull
out strength is POOR and requires helicoils. I eventually designed a two
piece inner (nickle iron) and outer (magnesium) housing split at the intake
port centers which could be cast or CNC machined and the two pieces sealed
with "O" rings. The end housings can have the number of stationary gear
bolts doubled to clamp the two pieces together and a dozen bolts in
addition
to the temsion bolts will clamp the perimiter. The center housing doesn't
lend itself to center bolts but it was possible to have a male extension on
one side fit into a female recess on the second half with radial (3)
sealing
rings. These center rings must separate hot oil and coolant from each other
as well as prevent external leaks.( A lot of oil is returned through the
center housing slot.)
     After working out how to build the two piece housings it turns out
that making both pieces out of aluminum is lighter. The alloy eventually
settled on was the one used by briggs and stratton in their lawn mower
cylinder blocks (and also the chevrolet vega) which is a 390 alloy with 13
to 17 % silicon carbide nodules distributed through the casting.  This
provides a long term wear surface that can't flake off and it is one of the
stronger aluminum alloys.
    The detonation gun coating is basically a pulsed plasma spray using a
series of small explosions to accelerate  the coating into the base
material.  The ceramic coatings and Nikasil process contain silicon carbide
powder as the wear inhibitor in a matrix of an inert material. These are
also coatings and ALL have to some degree failures in staying attached and
chipping. In a cylindrical bore these coatings are fairly successful.
however thermal expansion on flat surfaces seem to be a problem.  Good
enough for a dirt bike but not on my aircraft.
      Another concern is how much weight reduction can be achieved.
Weighing the stock irons and adjusting for the density of aluminum gives a
false result.  The thickness of the rotor faces needs to reflect the 1/2
strength of the aluminum which increases weight. Also adding a few
reinforcing webs to direct coolant flow adds a little more.  Basically
everything that is not coolant is aluminum. The additional perimiter
fasteners and stationary gear bolts are another half pound.
   There must be goals other than weight to make this project worthwile.
For me it was the ability to stack up a six rotor engine although I now
intend to only use four rotors and the ability to add some material for
porting purposes and to straighten the 90 degree bend of the stock side
ports. A few things like parallel cooling  dynafocal mounting  and dry sump
oil provisions become design choices rather than add on patches. It is also
possible to draw up the wide three rotor housing using a stock front
stationary gear which would be a better return on weight reduction and
allow
stacking two three rotors rather than 3 two rotor modules.
      It might be helpful if you made up a list of things important to
your design goals, and how you would fabricate the parts once designed. For
me having a 28" swing by 91" long  LeBlonde lathe and 7.5 HP cincinatti
mill
in the shop, I designed around those. Basically The castings are faced to
thickness and polished in the lathe. O ring grooves in the gantry router
bought for making the casting patterns. Jeff has a design and has finished
parts of a 2 piece aluminum end housing which incorporates an iron wear
surface
with the edge located between the housing water "O" rings.  It has  of
course the fit and finish we have come to expect from Jeff. I am not sure
how many rotors this is for but it may include a one rotor version.
     Current work is lost due to windows problems but I have a few JPEG
images of the final design of the engine parts I can attach.

george grimes

Excellent!
You guys are way ahead of me as usual.
I have worked on the early porshce 911 magnesium engines for 15
years or so and am familiar with the problems that caused
porsche to shelve the mag case idea permanently. The mag cases
literally twisted when they started building them over
2.4liters. The rear main bolts crushed the bore on many of the
engines.

For a racing engine I could get away with it, but not for
aircraft. The loads on the rotary are so different than on an H6
engine that it seemed a much easier task to accomplish with mag.
The threaded housings are really what kill the weight savings
the most compared to alum. Threaded inserts would have to be
cast into the mag and have a large shoulder on them. At this
point we have lost the game, and it's back to aluminum.


 George, I will be in touch with you in the future
when I have more time to sort this issue.

Thank you for your input!

Ed Carver


Ed Carver,  The RX8 rotor housings have the water "O" ring grooves . There
is a generation of RX7 housings which also have the grooves in the rotor
housings. If you use either of these rotor housings your end housings can be
cast then faced off in a metal lathe to finished thickness. The rotor face
still needs lapped and a shop  Bryan Smith deals with  near Kanapolis
specializes in this for street racer RX7 engines for a reasonable price.
Brian is the guy that bought the extensive collection of 3 rotor  parts that
Bruce Turentine had collected. Bruce is specializing in two rotor engines
and focusing on firewall forward packages with some specialty end housings
and motor mounts.

      You might still use magnesium for your intermediate housing because
there are no major threaded holes in this part. Tension bolts just pass
through. The oil pan flange has minimal load and could be a through bolt as
part of the design. There are tapped bracket holes on top which apparently
deal with stabalizing the stock intake system and are not used. Paul has
advocated locating the third mount attachment off these holes on top of the
intermediate housing. An extension tab with the mount bolting through it
front to back would keep this possible.

    I suppose you will have to cast any magnesium parts (vaccum assist) and
finish the joint between the parts first. Then bond them before machining
the final thickness. This way only the final face machined has to be
perfectly parallel to the opposite side.

    In my case I am doing the two layer housings in order to have access
for CNC machining of the ports on a 3 axis mill. There is a cast in web on
the housings which blocks the cast in  intake ports if I am doing a "P" port
version because it seems like a good idea to be able to build either
configuration. If you are not doing side intake and exhaust ports then one
piece housings cast with a baked core would probably be better. You still
need a CNC machined pattern but only done once, the foundary does the
copies.

george grimes