Building the JU-02 gas producer

[racket]

Hi Anders

You'll need to make sure the shaft tunnel face is running perfectly true before bolting on the diffuser , but you can't skim the shaft tunnel face , thats why I suggested turning up something that the diffuser can be bolted to , even a single large bolt through the diffuser bore will hold it firm enough against the jig to check the diffuser vane running trueness.

We have to be extremely careful about compounding up of small deviations between the parts, especially as the shaft tunnel face is only half the diameter of the wheel , its always given me a hard time making sure the front and back of my diffuser were parallel, its one of the jobs that has to be "right" , and with my crappy old worn out lathe , not the easiest to achieve :-(

Looking forward to hearing how it checks out.

Cheers
John

[elventu]

Being a lathe machinist since about 15years...
Some suggestions.

Even with a old lathe, there is a simple way to fit on a part precisely.
First of all, check bearings.
Obviously you need a tight chuck line, not too tight, just the right pre-load on the bearings (if conical roller one).

Clean with compressed air and solvent the chuck.
It has to be free from chips and dirt.

Put on the chuck a bar, do a 60° pilot hole on it for the tailstock, then put the tailstock with a determined load (ypu can load it by hand if you have sensibility, or simple put a weight on the handle to have some "precision" on the load.

Now you can control the chuck off-center and its "planarity" with a dial (comparator?maybe this is the right term, in italian we call it "comparatore").

Write on the chuck the higher and lower point.

After that, you have to machine the bar you already put on the lathe to match precisely the hole where you need to put the part you want to measure, in this case, you have to copy the shaft size, leaving a small face to the rear part to push the housing on.
After that, another control to the chuck and for concentricity of the machined part have to be done.

Then, move the tailstock and put the part you have to check on the shaft, then re-position the tailstock in the same position with the same load.
Re-check the chuck with the dial, the measurements should be the same.
If not, try to correct it with a copper hammer (or similar, like a steel hammer with some alu scrap between it and the chuck).
Light hammering allows the parts to return in the same position.

Then, with the dial you can control what you need.

It is more simple doing it than explain, expecially when I try to explain with my english.

I use this system at job when I have to test machined parts that have other machining after the one I do on the lathe, or for the parts I have to machine after thermal treatment, it cause some warping, we work some tempered parts with CBN or ceramic tools, for example steel tempered rollers or pulleys for steel cables, then I have to center the external part (when I haven't to machine it after temper) to work a small clamping diameter and face, then I machine the hole clamping the part on the surface I worked after the thermal treatment.

Except in cases of severe warping (it happens), the parts can be machined to 0.02mm or less tolerances.

The same occours when a tube-shaped alu part are milled after turning.
Alu (some alloys better, some worst) is a shitty material to be machined precisely.

Any hole drilled on the side can be the cause of a warping, maybe hours after the machining.
I have had some trouble in the past with 6082 (anticorodal) and expecially 7075 (ergal) making parts for a factory that make cigarette packing machine.

Parts that was precisely machined in the 0.01mm, with some milling, threaded holes etc, begin to warp after one day or more.
Parts we checked with precise instrument like CMM and after the control were perfect, after a weekend setting in the factory were warped up to 0.2mm in some case, some pieces from the same bar have warped in different ways.

This is one of the reasons because some time ago I asked why you build the shaft tunnel in alu alloy instead of steel.
I have bad experience with alu, and in my opinion it isn't the best material for that job.

No doubt that some of the perfectly running engines have alu shaft tunnels, and no doubt about the work done on them, I reporting only my experience about similar parts I machined for other purpose.

I wrote a long and maybe repetitive post... I hope you can forgive me, I have some issues trying to think in english 😂

[Johansson]

Hi guys!

I´ll check the shaft tunnel face first and if it is running true I´ll bolt the diffusor to it, if not I will make up a jig. Thanks Elventu for the detailed explanation about the lathe!

Anyway, I might have found something. The oil drain pipe is bolted to the backside of the diffusor and when I checked the area that had rubbed the compressor was exactly where the drain pipe is.

A theory is that the drain pipe is bending the diffusor from different heat expansions between the parts, I can skip bolting the two together when assembling the engine so they are free to expand as they want.



On a side note, we had a wonderful day today. -26°C and bright sun all day long!



Cheers!
/Anders

[turboron]

Anders, thanks for sharing your beautiful day. I lived in upstate New York for a few years near Lake Erie. The had a great expression for the hard weather. It is "We have 10 months of winter and 2 months of tough sledding". LOL

Thanks, Ron

[racket]

Hi Anders

LOL..............minus 26 with zero humidity and sunshine , now that looks like a nice crisp day , I just woke to +25 degrees with 90% relative humidity at 5 am , its going to be a hot sticky day, something half way between us would be nice :-)

Interesting that your rub is adjacent to the plumbing , yep freedom to expand is a necessity with our engines , thats why I have a slip joint at the rear so that the internals can move independant to the outer can which is only fixed at one end ............also I've been a tad concerned about your thick rear wall plate and the possibilities of it "dishing" because of heating from the inside causing it to want to pull the outer can rearwards.

The stress relieving slots you cut in the diffuser rear wall webbing would allow flexing , is there any way they could be "filled" ??

Just to be sure the engine is assembled as its designed to be it might pay to check the wheel to shroud clearance prior to finally bolting down the front cover , ideally having the front cover bolt on last would be best .

I'm sure you'll find the cause of the rub :-)

Cheers
John

[Johansson]

Hi John,

We were out most of the day shopping the last stuff for Christmas, the cold bit the nose for sure.

It would be interesting to know what is going on inside the engine when it heats up and the pressure starts to rise, there are most likely many forces at play pushing and pulling the engine components.

I could weld the slots I cut in the diffuser wall, but I´d risk that the diffuser distorts a bit from the welding.

If the NGV plate distorts from the heat it would pull the can backwards and possibly bend the compressor cover, no way of knowing unfortunately.

But, since the engine was running with comp clearance all way up to 70.000rpm all I really need to do is to make sure that the clearance is even all around the compressor and make it a little bit larger at the tightest spot to give it some more space at full throttle.

Cheers!
/Anders

[racket]

Hi Anders

When you did the original machining and checking of the wheel shroud clearances did you notice any inconsistencies between clearances ?

I'm about ready to put the 12/118 back together after picking up the balanced wheel on Friday , I might redo my wheel clearance checks to see if anything has changed since I first made it , theres been plenty of "cycles" for things to move , so might be worth spending an hour or so to check , I'll post some picks of any "oddities" :-)

Cheers
John

[Johansson]

Hi John,

Not really, I spent a fair bit of time trying to get it as good as possible. I noticed a while ago that there seemed to be a little less radial clearance on one side of the compressor inducer, might be something that has settled in during the runs.

I think that a milled billet diffuser like Andys would be much better than a home cast one, much less risk of it warping when it has been milled from a solid chunk.

I´ll figure it out.

Cheers!
/Anders

[elventu]

This night is the 2nd one with temperature below zero here, in the middle of central/northern adriatic coast of Italy.

I'm used to cold weather, I lived for 36years in the alps, but my girlfriend and my friends from here are more used to "wet cold", like 3/10°C with high humidity...
I'm just have return home with my dog after a dinner in a restaurant with them, they run to home due of the "cold" weather.
On the other hand I'm killed by the wet cold they are used to, my bones ask me for hours in front of the wood stove I had in the alps after some hours in that conditions... 😶

In order to limit distortions of the parts, can't you add some steel reinforcements bolted to the plate?

Not the best way to do it, but maybe it can do something to make the plate more rigid...?


If someone is interested, I know that at my job they are used to work an aluminum alloy that don't warp for a customer that produces molds for plastic.
I can't remember what alloy is, but I know that it is very good to machine and it have a high tensile strenght.
Monday I ask for informations.

[turboron]

elventi, I agree on the cold wet weather. I was in Copenhagen, Denmark once in the winter. We took a walk by the harbor after I meal. After 5 minutes my teeth were chatter so bad I could not talk. Give me cold and dry anytime.

Thanks, Ron

[elventu]

Just a weather news.
It is snowing, here...😂
And nothing happens in the alps, I just asked to my mother.

Bad news for my collegues that want to go skying during the christmas holydays 😂
Tomorrow, handbrakes.

[finiteparts]

Hi Anders,

I finally located that paper, but unfortunately, I can't find a online version to share. It is by Borg Warner and was presented at the 9th International Conference on Turbochargers and Turbocharging back in 2010...

"Compressor Wheel Retention" K.Shoghi, M.Becker, T.Ducker, D. Metz and C. Oberste-Brandenburg (Borg Warner Turbo Systems),IMechE

I have started to take the information from the paper and write a quick program to solve for the required assembly torque, but free time around the Christmas season is sort of hard to find. For now I will just step trough the content of the paper and what they term as important for compressor retention...and as it turns out, a simple program is quickly getting complicated in order to capture the physics accurately.

They begin by stating that there is a minimum clamp load required to prevent slipping of the compressor and other parts, and a maximum clamp that the parts are capable of absorbing without taking on any permanent deformation. They work to define an assembly torque that will provide a clamp between the minimum and maximum values over the entire turbocharger operating range.

The minimum torque is set by the combination of the thermodynamic torque and the acceleration/deceleration torque. The thermodynamic torque is imposed by the aerodynamic load applied on the compressor as it provides the required pressure rise to the system. Since it takes an unbalanced torque to accelerate or decelerate the rotor, this must also be included in the calculations in order to make sure that the impeller does not slip when the turbine provides additional torque to speed up the rotor.

The maximum torque that can be applied is that which would cause the parts that are in the clamped assembly to take on a permanent deformation. To this, the 0.2% yield stress of each parts has to be defined and the operating conditions where the heaviest loads are imposed have to be understood. It is interesting that they claim that the most severe loading is imposed after shutdown when the turbine heat soaks through the shaft to the impeller causing it to grow faster than the steel shaft and impose a thermal driven clamping pressure. This pressure can deform the aluminum contact surface and lead to a loss of clamp after it has cooled down. Additional effects that are competing during operation include the axial shrinkage of the compressor wheel, the reduction in axial length of the parts due to embedding and then the change in axial length of each of the parts due to the thermal states that they are in at each operating point.

One of the hopes I had was that they would include some of the relevant variables that would be hard to come up with and we are in luck. The coefficient of friction that they found to best match the torque test data was given as mu=0.15. The include some other parameters, but I will discuss them when I get the program completed and checked out.

A final point of interest. They discuss the effects of improper clamping loads on the turbocharger operation. Excessive or low clamping can yield changes in rotor balance due to the movement of the part relative to the rotational center. Small changes as low as 1 to 3 um are shown to impact the balance and thus the noise, vibration and harshness. They show an interesting shot of a impeller that experienced an overspeed. The effect of the reducing axial length of the impeller due to the overspeed allowed the self tightening nut to turn further on the shaft and once the shaft speed came back down, the impeller tried to return to it's original axial length but due to the tighter shaft nut, the interface with the seal spacer was subjected to a higher than allowed contact stress and created a permanent deformation imprint of the spacer in the impeller.

All in all, it is a very interesting paper and I am hoping to have a working program soon if I can get some free time.

- Chris

[Johansson]

Elventu: I start to think this is more a case of misaligned parts than anything flexing inside the engine, so I will center the cover perfectly around the compressor with dowels and set the clearance. I think that adding metal brackets might only bring other problems.

Chris: That was very interesting reading! Shows that there is much more thought behind the designs than one can think. The result from Johns experiment showed that there should be no problem with the compressor clamping when torqued to specs, so I´ll sort the comp clearance out and assemble the rotor by the book.

The X846 compressor John sent me arrived today, after the holidays I will bring it to my local turbo dealer so he can balance it with my spare turbine wheel. I have ordered a spare compressor and a bearing set from him as well that will be balanced with the turbine wheel I used when I dropped the compressor. Then I have a complete spare rotor ready for when I start building JU-03, or in the worst case scenario if the JU-02 rotor gets damaged again.

Cheers!
/Anders

[monty]

Dec 13, 2018 16:06:34 GMT -5 racket said:

Hi Anders

Looks like Monty doesn't owe you a beer :-(


Cheers
John

Well, just to make sure I have my bases covered....... I drank a nice Christmas ale for me.

and then I had a delicious Imperial stout for Anders!

[Johansson]

Mmmm, imperial Stout!