Rotaxaru

[facthunter]

It will be doing a lot more than 10,000 RPM. Smaller = (much) faster. Nev

[RFguy]

yeah more like 100k.

right so that much heat gets sucked out, even though it is a flow through. I dont know too much about specifics of implementation of turbochargers.

I have seen the radiant heat of a exhaust side of a turbocharger melt a battery casing nearby.... (automotive).

 

anyway, when the engine is mounted and I clamp the cowls on, I'll make some choices on cooling. 

In the mean time, next post will be the engine mount interface plate. I am going to use the existing mount with the same Jabiru rubber suspensions onto the interface plate (12mm ally- same as the jab plate) 

 

[onetrack]

The additional exhaust heat from turbocharging is directly due to the substantially increased volume of air being pumped into the engine, resulting in more fuel being added, resulting in more power - which results in more exhaust heat.

 

A turbocharged engine under load has a vastly increased chance of burning valves, thanks to that heat. Not for nothing are heavy duty turbocharged diesel valves made of stellite and other heat-resisting metal alloys.

One of the interesting things about turbochargers is that increased EGT = more turbine speed = more air being forced in. This is because of the increased exhaust gas expansion with increased heat. 

 

The average turbocharger turbine blades usually run at 80,000 - 100,000RPM under full load. 

[RFguy]

Onetrack, apart from the additional fuel air charge - say 25%, (so 25% more heat) does in this case the back pressure that was never there before appreciably change the heat flow out of the cylinder ? 

 

Or is the turbo just soaking up the heat as the exhuast gases go past?

since the 6kW 912ULS to 30kW (915iS) is an appreciable jump for a 100 >> 145hp. increase

 

[facthunter]

You get nothing for nothing. In flight all this turbo and exhaust run red to white hot.  Without an intercooler the charge gases are hotter. The motor has more volume of combustibles. More heat for longer A more sustained push on the pistons. often oil cooling jets on the underside of the pistons and everything stronger. Nev

[onetrack]

Quote

RFguy - "does in this case the back pressure that was never there before appreciably change the heat flow out of the cylinder?" 

 

Oh yes, the heat flow out of the cylinder is increased by an appreciable amount. That's why an EGT gauge is an important part of the gauges on heavy duty diesels when you're running a turbo and operating at full engine loads. It's easy to melt valves if you keep the "pedal to the metal" with a heavy load. Probably not so crucial on cars, because the only time they get a real heat load, is long-distance, non-stop driving at high speeds in high ambient temperatures. But an EGT is a crucial gauge for aircraft engines, because they largely operate under heavy load.

 

Edited July 4, 2022 by onetrack

[RFguy]

(WRT previous comment ) and marine I presume.
I melted a piston once with Nitreous...

 

In my day of group G rally cars, we gleaned a bit more power from a good tuned extractor system, good inlet system, carefully port and polished heads, carefully set up carbs. cool air ducts , aggressive insane lift cams, and a little over bore..... But didnt touch turbochargers, too much heat under the hood a problem amongst all the non standard plumbing. I see those Turbos in  personal vehicles have come quite along way- they are oil AND water cooled now- water pump runs after oil stops to stop it charring up the oil in soak...

Edited July 4, 2022 by RFguy

[Ian]

Turbos make engines more efficient by recovering energy from the exhaust gas however

1. They increase the backpressure on the exhaust making valves hotter.
2. They compress the intake charge making it hotter PV=nRT
3. The turbine compressor blades transfer heat to the airflow as they're not 100% efficient
4. You are generating more energy in a smaller space. This means more heat to dump

Essentially it increases the heat load on the engine requiring more cooling. You can mitigate some of these issues with an intercooler however temperature remains an issue. Things like sodium filled valves also help with heat removal.

 

[RFguy]

Thanks for the commentds Ian,  onetrack.

In the case where you wanted normalization up to say DA  = 7000 (ρ = 0.78)  x1.28 is required. 3lbs boost + a bit for the compression heating.

it would seem easier in the rotax case to add the big bore kit (x 1.15) would get you closer.. and if  are refurbishing an old rotax, the big bore kit is a no brainer.

and Or a supercharger would appear to have less implementation undesirables.

hmm 25kPa pressure, 0.064m3/sec = 1.6kW  (2HP) input. (100% compressor efficiency)

assumes intercooler gets temp down and no pressure loss in the intercooler. add a bit more boost to counter the boost pressure temp 

Call it 70% efficiency for a small roots blower at low pressure ratio (x1.3)  (AMR500 etc)  = 3.2kW input, 4.2HP. sounds reasonable. charts say 80% at PR = 1.3 theo.

Edited July 4, 2022 by RFguy

[Ian]

I'd actually say that a supercharger is less desirable than a turbo for airplanes except for the back-pressure issue and you have some new issues. Turbo's operates very effectively at the reasonably constant loads that planes have compared to cars. The main reason that turbos have a reputation for more heat is simply because they can generate higher levels of boost and they run hot themselves.

 

I didn't mean to put you off, if you're aware of the issues you cater to them in your design and build.

More capacity is probably the simplest path to more power, however more power also increases heat load and it doesn't get you normalization at altitude.

Rotax has released sodium filled valves which might be a reasonable way to extract a bit more heat.

Also remember that as ambient density decreases, the required turbo size increases so size the turbo for your cruise altitude not sea level.

 

 

[facthunter]

If your motor already has cooling issues don't even consider supercharging it. Most Merlin installations have ground cooling concerns. You can't linger longer than a minimum time. Sodium cooled valves cost a bomb and weaken the stems. I doubt whether small diameter stems do that much to transfer the heat..  Nev

[RFguy]

Remember we'll talking normalizing here. But that does include a high airfield int he middle of summer..

It's assumed you can get the heat out so there are NO new heat issues and temps are at the bottom of the green range and held there.

In Lycoming sized valves, non sodium filled : 25% stem, 75 % seat  (heat transfer)/ Lycoming sodium filled  up to 50% goes out the stem...

Rotax and Jabiru have 7mm dia stems. yeah not much space in the middle for the sodium, but 'it's all relative'.

Ian points out Sodium valves for the rotax  were introduced for the 915.

[facthunter]

The effect on the PLANES performance at the high (and hot) airfield  (Density altitude) will bother you as much as the engines welfare. Nev

[Ian]

2 minutes ago, facthunter said:

If your motor already has cooling issues don't even consider supercharging it.

Good advice. However if you can increase cooling efficacy it might be fun 😉 but more likely lots of heartache. The early turbo lycomings and continentals had lots of issues, and they are still prone to heat related issues. It's hard to get right.

Does anyone know what the sodium filled valves for the rotax cost?

https://www.rotax-owner.com/en/support-topmenu/40-uncategorised/658-si914030-914

[facthunter]

No doubt crazy figures. With the high temps why wouldn't there be a tendency for the sodium and the valve material to alloy as mercury does?  Sodium is a very reactive metal.. Nev

[RFguy]

When Jabiru went from ? to Nimonic valves, they went from $25 to $75. 

I beleive mercury was used for stems before sodium,

fortunately more cooling is easy with a water cooled engine. just add  bigger coolers and know your pressure differentials required and what you have.  Rotax still needs *some* airflow on bores though. 

 

https://www.cps-parts.com/cps/catalog/pdf/18.pdf

915

11 854-101 INTAKE VALVE 38 MM 4 $187.82
12 854-113 EXHAUST VALVE 32 MM 4 CALL

 

912 $187 / $304

 

An Aeromomentum AM13 is looking good 

 

 

Edited July 5, 2022 by RFguy

[facthunter]

You don't need anything fancy for the inlets. Nev

[Blueadventures]

Hi RF  do you have performance data from others that have a Rotax in a 230 Jab and then how that compares with the 3300's.

 

As you may get slightly better speed with the E-prop, keeping the wings and fuse well polished and also any drag cleanup you can do.  Such may save the turbo fit.

[RFguy]

other's performance data is anecdotal as similar.

My take on it : 

1) The 4 blade E-props should improve the initial TO roll substantially.

2) Expecting top speed to be down . (75% in the 3300 is 90hp) (75% in the rotax is 75hp) . But prop efficiency  (expecting a few percent) will be up slightly so the difference will less than that on paper.

3) Expecting to notice that HP difference at > 6000' when climbing between altitudes.  (in an RV you can really zip up to the next level  to 'have a look ') 

4) Can run the rotax at 85% (85hp) , given that I will keep all the temps in low green, clean oil ,  but higher load etc gearbox. 

5) Big bore rotax kit (115HP) is the big equaliser for cruise . 

6) Prop can gain quite a bit on efficiency with slower RPM.

then :

7) will clean up airflow on mains , that will improve a bit

8 ) substantial reduction in cowling lip drag  (because I wont need it) and optimizing exits  will be worth a couple of kts also.

 

9 ) clean up doors fit. 

10) clean up airflow on landing light, front cowl optimization.

Edited July 5, 2022 by RFguy

[Blueadventures]

6 minutes ago, RFguy said:

other's performance data is anecdotal as similar.

My take on it : 

1) The 4 blade E-props should improve the initial TO roll substantially.

2) Expecting top speed to be down . (75% in the 3300 is 90hp) (75% in the rotax is 75hp) . But prop efficiency  (expecting a few percent) will be up slightly so the difference will less than that on paper.

3) Expecting to notice that HP difference at > 6000' when climbing between altitudes.  (in an RV you can really zip up to the next level  to 'have a look ') 

4) Can run the rotax at 85% (85hp) , given that I will keep all the temps in low green, clean oil ,  but higher load etc gearbox. 

5) Big bore rotax kit (115HP) is the big equaliser for cruise . 

6) Prop can gain quite a bit on efficiency with slower RPM.

then :

7) will clean up airflow on mains , that will improve a bit

8 ) substantial reduction in cowling lip drag  (because I wont need it) and optimizing exits  will be worth a couple of kts also.

 

9 ) clean up doors fit. 

10) clean up airflow on landing light, front cowl optimization.

Have you considered polishing the wing the glider guys do it for added performance; its not just a matter of polishing the surface as you can effect the profile shape.  They correct the profile shape if possible.  One outfit is  https://maddogcomposites.com.au/services/glider-repairs/   may be of interest to you.  I missed an opportunity years ago to learn about the correct way to polish fibreglass glider wings.  There is a lot to be learnt from the glider performance ways.

[RFguy]

I think I'll just burn more gasoline instead.

[RFguy]

Update : OIl cooler spec is 10kW or 25sq" and water is 28kW 77 sq" (rotax 912ULS book)

 

Edited July 5, 2022 by RFguy

[RFguy]

For the drawing in fine, see the bitmap attached.

The jab mount interface with the standard Jab mount, and has the standard compliant donuts doesnt need any further thought- that is per jabiru design.

 

But the interface to the ring mount needs some thought - Calcs done for 10g shock load

Is 10g enough ?

Shear and overturning moment Loads on the bolts are OK at that. AND

- the Jab mount- the rubber donuts are compliant so that the transient- peak shock load at the  ring mount will not all be imparted on the ring mount-plate bolts . 

 

However- the M8 ring mount bolt running through the ring mount and aluminium plate - on the ally plate- it is within the min 100 Mpa proof stress of the plate , but  load there requires those holes to be tight fit to maximize the load area .

 

Spacer between ring mount and ally interface plate ideally needs to be a milled ally block with the ring mount radiused into it for a tube  snug mate or a tubing saddle washer used.

Preload tension in those bolts needs some thought, if the spacer is non compressible,  regular cyclic loads should be within the preload. 

If the spacer is compliant, the bolts will see more shear and overturning loads. Does the spacer need to be pinned to the ally plate ? Spacer should be sized such that the face clamp forces do that work. 

 

conclusion - Overall, the peak loads are well within the bolt's and interface plate capability assuming the interface plate was fixed in space. Because the existing Jabiru shock mount is present , the full peak loads will not be present on the ring--interface plate interface .   

If the interface bolts were tightened to 16Nm, (on a 5.6 bolt) there would be approx 10kN preload (~ 80% of proof)  for k=0.2 , which would be plenty. actually likely to be 2x more than necessary. A 16mm dia washer on the ally plate would generate 69Mpa on the ally plate  - need to check if that is a long-term stress problem even though way below the max. 

 

The non compressible spacer is an easily calculable solution- I'm left with nearly pure tension and sheer loads .

A compliant, compressible spacer is a different story.

I will have to do the calcs on the bending stiffness of the interface bolts IF there is compilance in the spacer. and this is bending on a preloaded bolt, just to complicate things. which likely means that expecting peak bending stresses + required preload should be < ultimate stress. 

 

I will need to do some research on the different behaviour (elastisicity ) of the different bolt numbers/materials.  = like ASTM307 is quite elastic compared to A574

 

What is my analysis like - Engineering comments please ? (useful comments only please, not commentary) 

 

 

 

jabinf1.bmp

Edited July 17, 2022 by RFguy

[RFguy]

post script - bending moment calcs on the interface bolts show that the 40mm spacer MUST be non compressible , non compliant,  as the 10g overturning moment peak shock loads are wayyyyyyy  above any M8 bolt capability.

[Thruster88]

12 minutes ago, RFguy said:

post script - bending moment calcs on the interface bolts show that the 40mm spacer MUST be non compressible , non compliant,  as the 10g overturning moment peak shock loads are wayyyyyyy  above any M8 bolt capability.

You could get the adaptor milled from 50mm plate with nice radius curves on the four integrated spacers.  M8 12.9 grade socket head cap screws?