Rotaxaru

[RFguy]

8740 which is what most AN bolts are made from has a pretty nice elasticity =  22%  at 50mm....

the harder stronger they are generally the less desirably elasticity properties (that we want)

12.9 is about 2x overkill on tension yield and has only 8% on ellongation.  10.9 about the same. but wow elongation limits for those two are over 1 giga Pascal !

we want a nice bendy bolt. strength is not the problem in the 8mm size. 4.6, 5.6 suitable.  wow the 8740 really is the biz on properties. nice.

https://fullerfasteners.com/tech/mechanical-properties/

If you had to describe a 8740 (like AN5 etc) in  metric head terms it would be something like a 7.6,  In metric world, I want something like a 5.6 (500Mpa tensile, 6/10ths of that at yield) 

I could use an AN5, shaft is 7.93mm, root diameter is ~ 7.  for metric is 8mm / ~ 6.4mm ! so..... might as well use AN bolts . Ring mount tube hole is 8.5mm ID. a bit big, but I am relying on clamping to hold this all together and the tube through the ring mount will be a column supported at BOTH ends so s SUPER strong. 

AN5 torqued to low end 11.24Nm would be approx (k=0.2) would be ~ 8kN tension (about 50% of limit for threaded part)  which  still heaps for the required preload

- IE low preload provides for some margin required for shear stresses and bending. 

 

 

 

Edited July 17, 2022 by RFguy

[facthunter]

10G is probably adequate. Above that what are you achieving. It's got to let go somewhere.. the (3) bolts holding most jet engines in aren't very generous. Engines detaching aren't that rare when things go bad.  Nev

[RFguy]

Thanks Nev- you can see my addendum above,  low ish torque (11Nm)  on the AN5 bolt is still  above required preload tension  and provides the bolt plenty of headroom for shear and bending loads. 

Yeah if the aircraft is good for 6g... and the rubber mounts to the airframe will take up quite a bit. These AN bolts (8740 mostly) really have very nice properties.

 

 

 

 

Edited July 17, 2022 by RFguy

[facthunter]

10 is also the factor for the seatbelt mount if I recall correctly. I reckon it's spot on. (If that means anything.)  Nev

[RFguy]

Oh 10g for seatbelts. yeah they'd want to hold on so you dont break your neck when the plane gets a 6g smack mid air.

The thing  is with the AN5, a single long bolt that is going to be an AN5-34 to get through the stack !

That's long, and elasticity will come into it. So I want to stay as low as I can (and meet preload, and required face clamping forces) .

10kN on the AN5 bolt 90mm long of unthreaded is an elongation 0.09mm . SO I can measure stretch to see what's going on. 

 

An alternative would be threading the spacer block 5/16-24 each side , and  have bolts into it . any preferences out there ? hmm will need to do ally thread strength calcs and have anti nasty in the threads which complicates torque translation into tension.  

A bolt sounds less to go wrong. 

 

 

 

 

 

 

Edited July 17, 2022 by RFguy

[facthunter]

A spacer certainly complicates your shear calcs.. Do a destruct test on a simulated join.. Nev

[RFguy]

the spacer is ok, with bolt all the way through  IF the there is no movement  of course- IE clamping forces exceed shear forces.

I've done the numbers of the shear forces and we're OK with 10g all on a single bolt... (and we have four) .

AT the end of the day - the rubber jabiru donut mounts are going to soften the peaks I think. 

Edited July 17, 2022 by RFguy

[facthunter]

It's easy to want to overdo it.. A lot of engine mounts look delicate until you check them out. Nev

[RFguy]

So- the problem will be with the spacer.

at the aluminium plate end- that will be preload tension will exceed the requiring clamping force to stop it slipping. 

but at the round mount end, to ensure the bolt doesnt see any bending, that fit from the spacer into the mount and thru tube would need to be perfect-  and it wont be.

 

This means I'll need to thread into the spacer at each end. This way, even with a less-than-perfect mate between the ring mount tube and the spacer, any bending load would be predominantly shear because the bending  (cantilever distance) will be extremely small. and the spacer block can be counterbored oh so slightly to take the thru tube. Still, not super. 

 

Combined rigidity forgotten -

the ring mount is damn rigid, and it is braced to the engine.... and  the 12mm interface plate is damn rigid. plenty above the max dynamic forces at 10g !

 

IE the top bolts wont be straight when the bottom bolts are bent.... it's all one assembly, braced. so the interface bolts will see the load together, never individually,  

 

However, there will still be the matter of the spacer to the ring mount joint not being rigid. So that WILL be a bending moment. That will need some thinking about. 

 

Oh, I forgot something - the prop might pull 200kg on the front , so we need to add another 2kN total tensile to the preload requirement . To date the only loads are static weight of the engine and associated bending moments at 10g.

 

 

 

 

 

Edited July 17, 2022 by RFguy

[facthunter]

That weight consideration is minor compared with the dynamic loading of a blade throw. With most of them the engine comes out and probably a good thing that it does. There's occasions where something's got to give. Nev

[RFguy]

Because this is braced in 2 dimensions, translated forces will all be tension or shear, no bending. 

 

1) The ally  spacer block onto the ally plate can be assumed one of the same, no movement there

2) the spacer and plate are drilled tight clearance onto the shaft

3) Shim the 7.9mm AN5 shaft up to the 8.5mm ID ring mount thru tube to eliminate movement inside the oversize thru tube

4) counterbore the spacer to accept the thru tube and provide a butt to the block

 

Prop torque is bugger all. (until as Nev says, you throw a blade or get a prop strike) 

So life is good. OK I think I have done enough analysis. In practice there will be some play andimperfect fit and the bolts are suitable for the calculated small bending load. 

If the joint becomes loose though, (IE tension comes off the system)  that is bad !

 

 

 

 

[onetrack]

RFguy - Check the bolts you plan to use, for accuracy of shank dimensions. Most fasteners have quite a wide variation in allowable shank finished dimensions.

[RFguy]

Hi OneTrack. thanks for the point.

This reminds me a bit now of another discussion I was reading about of the requirement of precise fitting of propellor flange bolts in and thru the flange.

and that specific high tolerance bolts were available. 

 

I also read that while the AN desig doesnt specify the material but does impose a minimum tensile of 125kpsi. somewhere around 6.X to 7.X

 

also in the AN paperwork, max recommended torque in shear load is 9.6Nm (tension of only about 3.6kN fpr k=0.18)  (compared to 15.8Nm in only tension)

reference : AC 43.13-1A

in normal operation and within airframe maxima, loads will be well within bolt limits.

 

[RFguy]

Received the four blade propeller for my 912 J230 from Eprops. 

very nice. Packed very well. Hub very solid  and weighs nothing. came with blade  covers and digital protractor tool and some good reading.  This is a quality product so far.

 

using phone camera there is some geometric distortion of the photo straight lines . Titanium leading edge, that will keep the grass down at YCWR.

 

Edited July 19, 2022 by RFguy

[BrendAn]

On 30/06/2022 at 11:41 AM, RFguy said:

Nev, balance is not too much of a problem. Balance looks OK on the provided CG locations - that will be OK In fact, I might have to add more weight to the front ! but that's OK, move ancilliaries around etc

..... I'm maining concerned with keeping the location of the prop as near as possible as it was .  If the CG is identical and prop location identical, that makes it easier to show airworthiness in that I didnt change those two items.

Certainly, this method is NOT applicable to Jabiru 4 cyl replacements- that needs a different mount to get the engine back, or a direct mount from firewall to rotax case lugs.

Am I missing something about the weight of the rotax compared to Jabiru. According to the spec sheets the 912 (80hp) ramp weight is only 1 kg more than a jab 2200 ramp weight. But everyone seems to think there the rotax is a lot heavier. The 3300 would be heavier than a 912 according to the data I read. 

[RFguy]

you're not missing anything if you read the specifications carefully of each engine.

depends on how people count accessories. yeah the 3300 is heavier but the rotax mounts further forward, some 240mm further,  and the CG will be about right.  once you add in the weight of the ring mount, couple of 2kg,  the 12mm adaptor plate couple of kg etc the radiator hanging out the front another couple of kg , it's not too dissimilar. 

[BrendAn]

8 minutes ago, RFguy said:

you're not missing anything if you read the specifications carefully of each engine.

depends on how people count accessories. yeah the 3300 is heavier but the rotax mounts further forward, some 240mm further,  and the CG will be about right.  once you add in the weight of the ring mount, couple of 2kg,  the 12mm adaptor plate couple of kg etc the radiator hanging out the front another couple of kg , it's not too dissimilar. 

Thanks Rf.  I thought I was reading right. I have an lsa55 jabirax. The engine mount is pretty close to the firewall and the oil tank is beside the fuel tank. It replaced the kfm engine which may have been lighter than a 2200.

[Marty_d]

RF, my build uses AN7-40 bolts from the mount to the ring mount.  From memory the AN7 size was just right for the holes in the ring mount.

[RFguy]

Hi Marty, AN7 ?

! they're massive. 

holes are 8.5mm ID.. so AN5 for me. 

[Marty_d]

Maybe mine had been bored out, although it looked new (bought it second hand).  I got the AN7's because that's what exactly fitted.

[RFguy]

After marking it up on a big sheet of plywood

now, 

mucking around in the CAD, the cyan is the superimposed engine mount tube outline. 

big holes are the jab mount interface. little orange ones are the rotax ring, and the crosshairs are ther thrustline. 

will cut it and them bolt engine to it and see how it looks. once the plywood mocks are correct I will get someone to cut the big plate....

tomorrow I will do a bit more of wher eI wil retain the plate and wher eit will be cut out.  view is looking in from the prop. (front) , looking at the firewall.

 

 

 

[Marty_d]

Have you tried 3D cad?  I used Sketchup for mine - easy to use and you can put your engine in place with the thrust offset and everything.

[BrendAn]

8 hours ago, RFguy said:

After marking it up on a big sheet of plywood

now, 

mucking around in the CAD, the cyan is the superimposed engine mount tube outline. 

big holes are the jab mount interface. little orange ones are the rotax ring, and the crosshairs are ther thrustline. 

will cut it and them bolt engine to it and see how it looks. once the plywood mocks are correct I will get someone to cut the big plate....

tomorrow I will do a bit more of wher eI wil retain the plate and wher eit will be cut out.  view is looking in from the prop. (front) , looking at the firewall.

 

 

 

Can't you buy a ready made engine mount for rotax to Jabiru.  It would be lighter than using adaptor plates. 

[RFguy]

"Can't you buy a ready made engine mount for rotax to Jabiru.  It would be lighter than using adaptor plates. "

No - you cant buy a mount, and also every Jabiru is slightly different.....and I dont want to save weight- I need to keep the aircraft as close as possible to the original book. I will add weight if I am low.

I have done the new CG calcs and we'll see how that comes out in a weighing and where I need to add the weight.

 

Marty- I do have 3d cad but this is a 2d requirement. I will just DXF export directly from my CAD application that I am using....

First cut and install will be with ply. we'll see how that looks.

 

[RossK]

RF, when you get around to getting the ally plate cut, water jet cutting gives a much better finish in that thickness.