TV94 Turboshaft Project

[ashpowers]

Hi Dennis,

LOL, no I didn't get the feeling that you were sticking up for them and I do appreciate your input! Thing is, I sent them a .STEP file which they imported into their CAM software to machine the part. There was no part programming in this case, just tool choice, feeds, speeds, lead in and lead out.... And they obviously did not drill the mounting holes with the CNC, which was a big mistake....

[ashpowers]

Had to go to a water cooled processor system - air cooling just wasn't keeping up.

Corsair H60 CPU Water Cooling System installed. Running at 4.5GHz, 1.23VCore, temp 44-48-45-43 C. Will have to run a prime95 test to make sure she will stay stable, but these cores have been run at much higher freq and vcore voltages by others. I'll prob just leave it alone as this is a smoking fast workstation. =)



What a cute little double-pass tube/fin radiator core! =)

[turbochris]

way nice overclock! ha ha don't let richard see this, he's an amd man he'll delete this post!

[ashpowers]

I almost went with the AMD FX octo-core setup but there is a ton of info online, benchmarks, etc, and even the 2600k i7 was outperforming it. The 2700k, if I am not mistaken, is basically just a "select" 2600k processor - allowing better overclocking. Either way, the reviews led me to my decision to go "i", and I feel this system has met and exceeded my expectations. =)

[Richard OConnell]

Nah, I love Intel for projects where I see more necessity, but at the time, you were trying to encourage me to buy an i5 over my AMD Phenom II 1100T. I'm still very sure that the AMD's performance far outweighed the steep price difference for its intel equivalent, even more so when compared to the high end i7's at the time ($1200). Plus the new generation 384-bit graphics cards seem to do a pretty good job of taking the load off of the processor for heavy tasks.

[ashpowers]

I was in IT for 14 years, starting in high school and fixing computers at a local computer company after school. I started with MSDOS 6.11, Win 3.1, Sound Blaster 16. That was an intel 386DX-16MHz, 2MB ram, 120MB HDD. I got "bit" by the technology back then and that laid a course for my life for over a decade. I mastered it. I have owned both AMD and intel products through the course of the years; afterall, they are the two largest competitors in that market. Sometimes AMD is ahead, sometimes Intel is ahead - its a constant race - a battle between the validity of Moore's law and the reality of the technology's progression. Sometimes I've wondered if they are both in kahoots with each other given the small but consistent progression of processing power evolution.

Reality is, there is no debate unless you just want to debate. Who has the better product? LOL, that almost depends on what day of the week you ask the question. =) However, being that AMD is the underdog in the industry, I root for them in all cases. Doesn't mean I'll buy their product that thursday I go seeking; it just means that I appreciate the weighted challenges AMD faces in the industry and the perseverance the company has shown over all these years. All things being equal, AMD has pretty consistently offered a better product for the money.... and they keep on at it. =)

[turbochris]

you guys would have had a good time watching me play whack-a-mole with a virus earlier this week. It infected any storage device attached via usb and crawled the network every 30 min. it was starting up as a service.......

I think an i5 would run games faster than an 1100t

[Richard OConnell]

Lol, I LOVE virus whack-a-mole. It gives you something to do other than installing software updates

[ashpowers]

Well, a little update here. After sending multiple pictures of the machining flaws and going into detail about why the piece is unacceptable, the machine shop has agreed to address the issues. They have requested that I send the part back to them. I told them that this part cannot be "repaired" and I will not accept it back under any condition. I presume they want the part back before they do anything so they can see the flaws themselves and perhaps to make sure I'm not just trying to get another part for free. Fair enough - they can have it back.

Another interesting note, from the images showing the measurements of the variation in mounting hole position, I was measuring in inches. Showing one set at 0.794" and one at .824". She replied back and told me the following:

1. Our engineer measured the size between two holes is: 8.8086mm, In your pictures, you mentioned two numbers, it is 0.794mm and 0.824.
As you know the tolerance of machining is +/- 0.05mm, you will see that all is the right dimensions.

LOL and ROFL! The measurement was taken in inches, not mm. There is 0.030" variation there, which is 0.762mm - over 3/4mm variation. If they had machined the holes to within 0.05mm, which is .002", all would be well. Point here is, they were targeting 0.002" as their machining tolerance, she stated it, and the part is nowhere near that.

So, once they get the part back and measure it themselves they will see I was truthful. Fingers are still crossed though..

[ashpowers]

I've been in contact with a company, Dyna Tech Engineering, who has a few software packages available for rotordynamic and rotorbearing analyses. I have been speaking with one of their reps over the past couple of weeks about one software package in particular: FEMRDYN.

www.dynatechengr.com/femrdyn.htm

Things are looking good to purchase a license for this software. It will allow me to fully model the rotor group and come up with a design to address the complications seen when using rolling element bearing systems in these engines.

One aspect of these rotorgroups I've seen as the culprit for a majority of the damaging orbit/whirl modes actually comes from the compressor end, not the turbine end. This is most likely because the stiffness of the shaft passing through the compressor wheel is reduced because of its smaller diameter. I have considered also using a bearing support at the front end of the shaft, in front of the compressor wheel.

This software will be able to resolve the question as to if that will be of benefit. Regardless of that result, I'm pretty confident that employing this type of FEA to the design will result in a system that will actually work. =)

.... should know probably by tomorrow if I'll be getting this software... may even have a copy before the day's end.

[Johansson]

You are definitively taking this ball bearing work seriously Ash, I guess it is well worth all the trouble if you end up with a system that works spot on directly instead of having to R&D your way through several sets of rotary assemblies before everything is sorted out.

It would be great to use ball bearings and get rid of the oil pumps, tank, regulator and all damn hoses but I´d rather keep them than go through a research project of this magnitude.

[dennis]

Hi Ash,
if your measurements in the pics were really in mm instead of inches, this would be some seriously incredible engineering
Oh and some seriously small fingers too! ;D

[ashpowers]

Hi Anders,

As you noted, the expense of these rotating bits aren't cheap. I could endure replacing several $100 T04 turbines and $50 T3 comp wheels in the smaller engine when the bearings grenaded but my disposable income hasn't grown proportionally with the change of scale from T04's to TV94s, LOL. With the greatly reduced price they are offering to sell me the rotordynamic software for ($4000 retail, $400 my cost), it is less than the cost of a rotating group and will allow me to properly design a rotorbearing group that will work. And of course, my findings will be shared here so my efforts will contribute to the group and likely be of benefit to others' projects.

I've gone through the "lets build it, fire it up, cross our fingers and hope for the best" approach before. Because I've set the bar so high with this project, that being with the requirements of engine reliability/longevity for use in an aircraft, I'm very motivated to use any and all tools available in the design of this engine. I've enjoyed the thrill of firing up a new gas turbine engine for the first time on multiple occasions and although I know it will be just as satisfying to do the same with this new engine, I'm also not in such a hurry to get to that point. Im at a point now where in order to experience that same level of "virgin" excitement again, I'll have to pull off an engineering challenge that results in a solid rotorbearing lifespan.

The "cheap" T04 project had a purpose; to build an engine literally "off-the-cuff" and see what areas scream for more attention. Rotorbearing failures were the biggest hurdle and I'm marching in with the big guns 'a blazing... =)

[ashpowers]

Allright, I'm back! Received the replacement NGV ring today and the mounting holes are all within 0.001" placement - perfect! I'm a little upset that they chamfered the ring edges again, but it is very minor and shouldn't affect the final performance. At this point though I will not be using that shop again. Just too many problems and oversights to continue working with them.

Having my local machinist source the plate I need to make the outer NGV ring and the exducer pre-swirler ring - he will be getting on that Monday and hopefully I'll have some material to start machining later next week to finally get the hot section parts fabricated.

I've also put in a purchase order for the rotordynamic/rotorbearing software - going to bite the bullet and spend the $400 to get my hands on it!

[Johansson]

Sweet!