Hot rear cylinder example Jabiru 3300A-2575 and plenums

[RFguy]

Hi . OK, I guess you are referring mainly to the HEAD SCREWS. (the through bolts are pretty cool) .... OK you like the nickel . Yeah I have a selection of anti seizes, mostly due to the prevalance of aluminium/SS hardware in radio masts, and I (now) have a selection of high temp anti seize pastes.   Galling of SS bolts on radio mast hardware is a nightmare, as you can imagine... having an M12 SS bolt galled on a bracket 100m above the ground....

 

OK on the shank inspection .

I can measure root diameter/ shank diameter. I will measure on the HEAD SCREWS  that came out. The leakage will be screw damage or head recession , or deformation due to high temperature differentials (likely)  between top and underneath..

cheers

 

Edited February 27, 2021 by RFguy

[Bruce Tuncks]

Onetrack, I like your use of neverseeze but I was surprised at only a 10% reduction in applied torque. I would have thought maybe 25% but this is just a wild guess.

[RFguy]

basically all bets of off precise tension versus torque  unless thread/face friction is well known. 
 

According to peer reviewed rigourous math / engineering evaluation (S Berger) , seems that the Jab values for torque are pretty right IF the theads and friction faces are of low friction... IE well lubed. IE someone did do the calcs, but didnt take into account variability of thread/nut face friction for subsequent or reuse..... With mfrs plating and surface treatment , the preload aceived on Jab's torque values may be close, but subsequent uses of the thru bolts will have varying plating and surface/.thread friction compared to new, clean, plated, worse case dry/ unprepared would lead to 50% of preload being acheived at specified torque (S.Berger).
 

I'm lookup up again what the thru bolt mfr recommends, some fancy ultra high pressure lube, but in discussion witht eh mfr, ordinary 100 on super clean threads will be close enough...

I will be measuring the elongation of the thru bolts to  get a known cold preload. 

[RFguy]

this is a good one 

Bolts lecture AL 2017.pdf (usyd.edu.au)

 

particularly section 21, preloading. and methods to determine : 
and I quote "It is suggested that a preload of 75% of proof stress be used for the typical reusable connection"

summary , in order of accuracy : 
torque wrenches : "realistically at best with an uncertainty of about  20% (2.5 sd) in the preload of the bolts if a torque wrench is used" (lube dependent) 
Turn angle :  The nuts are usually then torqued to a low preliminary level, followed by turning the nuts by a prescribed angle (~180 to 360) to provide the required bolt stem elongation.

Direct length measurements. Possibly the most accurate means to determine the bolt preload is by direct measurement of the bolt length before and while the nut is turned.
 

The ARP website has a useful tech section

The Official ARP Web Site | Technical Information (arp-bolts.com)

 

Seems that in 'installation' standard oil isnt very good, (much better than nothing)  compared to their fancy lube. (although that was at 18,000 psi and 120 ftlbs - rather high....) .

I'll get myself some of their fancy lube for the thru bolts - available locally in australia easily enough.... this way if I have a fight with Jabiru, I will have used the manufacturers recommended method and lubricant. 

 

 

Edited February 27, 2021 by RFguy

[onetrack]

That ARP website is the most informative, and most clearly-written technical article I have ever read, in relation to fastener design, manufacture, installation, and associated problems. Good find.

[facthunter]

There's real problems with just using  a tension figure. In the end it's a "stretch" figure you need to get a force without exceeding the elastic limit for the Bolt or stud. The pitch of the thread will convert turn angle to a stretch figure so a seated low torque followed by a certain specified angle is more accurate and consistent..  There are also some assembly situations were you don't want oil involved at that stage of  assembly.   Nev

[RFguy]

Well, from tension, you get stretch of course after knowing a bit about the bolt ..... and back...

 

The numbers for the head screws (different job)  will be interesting given that the aluminium goes plastic and  the yield strength falls rather quickly over 200C. The head screws do have ~ 18mm diameter load spreaders (they feel like 4130)  . Based on some initial calcs of minimum preload requirements for  the  dynamic pulsating load, I would think the book torque for those screws with 'super lubed' k=0.1 condition would actually be a bit high, and be a little close to the compressive strength at 200C ...At unprepared/dry the book numbers are  closer to my calcs.  Almost like the screws are too big to get the preload up enough without causing trouble.  hence the recession perhaps that occurs (guess). More calcs and better assumptions on unknowns required.

 

 

[facthunter]

It's going to scuff with heat cycles whatever you do. That's why the retensioning of those particular bolts is needed in the very hot area.. Nev

[RFguy]

a bang in a jab cylinder is about 44kN. . the head- bore interface area is 1523mm2, so minimum pressure against the cylinder  just to meet the bang load is going to be  44kN/1523mm2 = 29 Mpa which is pretty low for the 200-250 MPa aluminium .

or 7.3kN/screw

Tensile strength for those 5/16 /24 screws . dunno, guess..if its grade 8... probably 33kN.  so say a preload of 15kN is 2x the bang load per screw. oh yeah cant be too far out  (not examining elastic  behaviours etc ) ....  15kN/screwx6 = 90kN and the 1532mm2  = 58 MPa alright  OK so far. 

(assumes even spread of loads which is not true- that's where I would need to do some reading to understand the likely distributions and what MEs assume for irregular load distribution in this case) .

 
70% of grade 8 5/16-24 ASTM??? bolt tension yield is 27kN  in a supplier's book
That Jabiru  torque spec 34Nm.  very roughly at K=0.2, that will generate approx 21kN  VERY ROUGHLY  - around 63% of yield. alright . seems fine.

At K=0.1, lubed up nicely, perhaps half that tension.  . alright 21kN/screw  = 126kN total, or pressure on the faces of 126kN/1523mm2 =

=82MPa. OK then.  

*Yeah the Jab numbers seem OK for dry*   so what do we get recession ?

The aluminium temperature would have to go over 250-280C for the Ally to reduce in yield strength if it was a step function (but its not - must be some flow over time ?).

Presumably there is some long term cold flow that occurs, I dont know anything about  flow of ally. someone will. I will ask.

HOWEVER, if the tension on the  cylinder was not perfectly evenly distributed , then we could easily meet  max yield stress temperatures at around 220-240C (depending on the temper, material) 

Edited February 28, 2021 by RFguy

[RFguy]

A little more on these head screws - and the dry/unprepared versus lubed case that has been discussed. 

If lube is used, say nothing special like W100, reducing k from 0.2 to perhaps 0.15 increases tension approx 33%. (although the thread engagement is relatively short for those head screws that would affect that number).

33% pushes the static cold tensions  to 89% of screw yield and 109 MPa at the interface of the head and cylinder..... 

As the head material expands as it gets hot,  the tension will ramp up as it iwll be like extra turns on the screw. (calcs required) -  And the Aluminium at 250degC might be down to anywhere between 100 and 150 MPa yield strength depending on temper, grade etc....

 

Now , you have flow of the Aluminium which will hit some equilibrium, and when it cools down, the tension will be off and it will leak around the interface/deformation area.

So, I would go with the unprepared case, and/or some fancy washers.
 

[facthunter]

The head is hotter  in the problem area AND the rate of thermal expansion is Higher for the Aluminium alloy by far., so there is a high (local) shear load. The rest of the seal area is not an issue in most motors. Retensioning done carefully covers this. All engineering is a compromise. IF you had a gasket there  it would present another problem. A slight leak is preferred to a section of gasket missing and without a gasket you get better heat transfer across that face. Nev

[onetrack]

Fasteners are the most critical area in coping with combustion, rotation and heat cycle strains. The ARP website reveals just how much attention is needed to fastener design, materials and heat treatment.

I've always gone for fasteners of the finest manufacture and strength that I can find. Many fasteners are mass produced on a huge scale, and suffer from undesirable variations from specifications.

Caterpillar roll the threads on their fasteners in manufacturing. and this is a far superior method of installing thread forms, because the grain of the metal follows the thread form, and isn't cut, as in manufacturing utilising cut threads.

Then the fastener material must be of exceptional quality as well. Carefully heat-treated low alloy steels with all the highly desirable additives such as nickel, chrome, molydenum, vanadium and cobalt, produce superb performance results.

The finest fasteners I've ever found are the long (300mm) bolts retaining the heads and barrels on Deutz diesel air-cooled engines. Those cunning Germans know steel, and those Deutz retaining bolts are incredible, you can twist them like a liquorice, and they will spring back with no retained twist. They must contain a lot of molbydenum, vanadium and cobalt, they're better than tool steel.

The Deutz diesel used no head gasket, merely an alloy finned head with a steel finned barrel. The barrel was simply lapped to the sealing surface of the head, and leaks were non-existent if done according to the manufacturers instructions.

 

Edited March 1, 2021 by onetrack

[Bruce Tuncks]

Sorry for going off-topic, but I have to say how trading in an old Deutz for a new Massey Ferguson was the biggest mistake since we got the farm. The MF was a heap of #### compared to the Deutz.

[Bruce Tuncks]

Back on topic... After 5 hours on mogas, the compressions were checked and they varied from 65/80 to 72/80. There was no big reduction at 15 degrees either side of TDC.

A borescope look inside the cylinders still showed  beige crap lightly crusting the combustion chamber. I think this beige stuff is unscavanged lead bromide.

My plan is to repeat the leakdown in another 20 hours or so to see if it gets better or not.  

[RFguy]

so.. beige crap. yeah

perhaps gunk getting behind the seats and guides getting dirty etc preventing full valve closing (that was my engine).

 

Will be interested to how it changes

 

also, I am beginning to think there could be good information of hot and cold leakdowns. On head -bore interface recession : I suspect that if there has been head recession on the lower head screw region, that might show up when the engine cools down (and a gap emerges because the aluminium yield strength was exceeded due to high temperaturw+ high tension). Maybe. 

 

Bruce, I plan on running Decalin Runup fuel additive when I return to service on AVGAS.

Edited March 5, 2021 by RFguy

[kgwilson]

I can't really see how a few hours running Mogas is going to get rid of the deposits already there from hundreds of hours running on Avgas which are pretty hard from all accounts. I've only ever used Avgas when away somewhere & I couldn't get Mogas so my engine has never really had the opportunity to get lead based deposits.

[RFguy]

I agree KG.  After days of chipping off the stuff, I consider the beige gunk to be like BRICKWORK on the insides of your CC. It's tough.  I got a big slab off in a dish and hit it with the butane  torch and nothing much happened..... MAPgas torch burned it though....

[Bruce Tuncks]

A Jabiru at our club is showing improved leakdowns after about 100 hours of mogas. The oil is darker than usual at change time, so it may well be cleaning up stuff. 

Too early for me to tell I know. I might age-out before the engine cleans out.

[RFguy]

Maybe detergents etc  in the fuel. What precisely is your 95/98 source ? just high octane ULP or is it one of the 'improved' fuels with detergents and other additives ?

Is that just one plane, or is there trend ?

 

cleaned up all my valves to new yesterday arvo. 

there is quite alot of gunk on the backs of the valves next to the seal where gunk builds up and prevents good sealing , due to the gunk on the seat lower face.  Just got to clean that lower seat (IE on the seat, but just below the mating face area) need a small rotary brush for that . Important to clean and make smooth in that region because rough bits will catch and enhance buildup.

Cooler cylinders , and cylinders that has jammed rings have a layer of hard black carbon on the combustion chamber UNDER the avgas bege gunk.

 

 

 

[Bruce Tuncks]

I think its a trend that the fleet is going better than on avgas.

I didn't know there are different versions of mogas 98. I'll look into it. So far, I have just bought at service stations and only checked for the absence of ethanol.

But here's a question... it seems to me that Yenn is quite right to add a bit of 2 stroke oil to the fuel. How else is the very top of the cylinder going to get any oil? Yet if this was such a good idea, why is it so discouraged?

[RFguy]

it might come down the valve guides :-) 

I am in favour of flooding the rocker compartments with oil flow to assist cooling.  But I dont know about other adverse issues with too much oil getting into the guides.  AFAIKT, oil that gets in the guides has a very hard time (gets burned) 

[Bruce Tuncks]

I have been trying to find out about fuel additives. One article said that you can't just run a leaded-fuel engine on unleaded without risking the engine, and that you needed some substitute chemicals. That's not what I've heard about  Jabiru engines. Maybe they have a low enough compression ratio ( 8 to one, maybe 8.3 to 1 ) that detonation is not a problem with 98 mogas? They did say that it was detonation and not valve-seat recession that was the problem.

The only negative thing I found about the 2 stroke oil was that it didn't help the exhaust valve seats, on account of having been burned before getting there. I also read that some petrols already have oil added...  how would you know?

[RFguy]

MY understanding is that's old world knowledge on valve stem/guide, seat etc lube of older materials and incompatibility with MOGAS, and that ye-olde  engines  that have been in planes for 50 years likely have new guides, valves, seats etc by now....  different materials. 

IE new valve seats are hardened , valce materials are hadened etc and dont need the TEL.

Rotax 912ULS is 10.8 : 1 . not such a biggie .... plenty of resistance to knocking with PULP98.

remember there is MON and RON.

in australian PULP 98 has RON  of 98.  but that is not the same scale as for AVGAS  

read up about AKI. AKI for PULP98 is about 93 ?  LL100 AKI is about  105 .

 

I think Rotax low head temperature , bore, piston etc (is water cooled) low temps have something to do with a lack of knock problems with Rotax.   I wouldnt run PULP in my jabiru in summer without a fuel return  back to the tank or header tank from a fuel pump bleed...for fear of vapor lock concerns.  

BUT BUT that is all covered well elsewhere on this SITE 

Back to the topic. hot engines.

 

 

 

 

[Old Koreelah]

2 hours ago, RFguy said:

... I wouldnt run PULP in my jabiru in summer without a fuel return  back to the tank or header tank from a fuel pump bleed...for fear of vapor lock concerns.

As one who is changing to MoGas, that made my think of ways we could cool the fuel line to the engine. Maybe a cool-air tube like the ones running to the coils and fuel pump?

[kgwilson]

5 hours ago, Bruce Tuncks said:

I didn't know there are different versions of mogas 98.

Each of the Mogas Brands have their own recipe of aromatics added to the standard fuel to boost the octane rating to 98 RON and they give them names to entice you to buy them. Shell calls it V-Power, BP calls it Ultimate, Caltex calls it Vortex. Aromatics include Toluene, Benzene and Xylene. There are lots of others like naphthalene which is normally solid but dissolves when mixed with other aromatics and then there are polyaromatic and hydroaromatic compounds.

 

Avgas is different as it is based on Paraffin which is the flammable compound used in candle wax and requires TEL (tetraethyl lead) to be added to achieve its octane rating. It is more stable and lasts longer and has a higher Reid Vapour Pressure (higher boiling point) making it less susceptible to vapour lock at altitude.

Edited March 8, 2021 by kgwilson