Sign up Calendar Latest Topics Donate
 
 
 


Reply
  Author   Comment  
OLi

Sr. Member
Registered:
Posts: 1,915
Reply with quote  #1 

We happened to end up at a GT that had blown up due to to much hydrogen gas. Twin coaxial shaft 5000/7000RPM at full load. I rambled about it in another thread but it is worth it's own,
Some blades were replaced, combustion chambers etc. Coupling bolts were clean shear off etc.
Shaft ended up slightly longer (high speed) if I understood correct new blades were heavier and thus increased rotor weight I will verify and modify if wrong. Bearing positions regrown that reduce shaft diam slightly locally and otherwise it is running better then ever more efficient and good low vibration.
So after such event the critical speed increased 100-150 RPM..... They run slightly above critical, less now.
Unbalance sensitivity increased at least 50%.
I thought that was not what I expected unless some modal magic or moving bearing position that is quite possible.

So if this transient did change the mechanical properties of the shaft will it make it more brittle and increase risk of cracks? Or can you just shut your eyes cross your fingers and run?
I must say the more I think of it the more worried I get.


__________________
Good Vibrations since early 1950's, first patented vibrometer 1956 in the US.
http://www.vtab.se
Alex

Avatar / Picture

Sr. Member
Registered:
Posts: 204
Reply with quote  #2 
Close your eyes and keep the motor in your car running [biggrin]

How much is: slightly above critical?


OLi

Sr. Member
Registered:
Posts: 1,915
Reply with quote  #3 
I will check but previously somewhere 5-800 RPM I think and now 100 RPM closer " but it is quite sharp" but it made a significant change in unbalance sensitivity
so it is "under the influence" still adding and removing test weight was reasonable repeatable and only 3-3.5 mm/s 1xRPM but 10 gram gave increase to 4.5-5 mm/s...
It is like 8 hour flight away incl. swapping planes so it is far away enough.
I will suggest they monitor 2xRPM vector for the slim chance you get time to see any change in time but you never know a 600mm generator shaft took 2+ weeks
at 3000 RPM but this is thinner. They run it on waste gas so they may find a way to blow it up anyway.

__________________
Good Vibrations since early 1950's, first patented vibrometer 1956 in the US.
http://www.vtab.se
Alex

Avatar / Picture

Sr. Member
Registered:
Posts: 204
Reply with quote  #4 
So, a quick approx. calculcation says they are running 10% over the critical speed. In my opinion that is quite close and when you are saying about change in balancing sensitivity... I think it is dangerously close or walking on the edge of it. In this case I would take a Bode plot. It should tell you more precisely how close you really are.
OLi

Sr. Member
Registered:
Posts: 1,915
Reply with quote  #5 

We tried but the tacho wouldn't let us. It seems to move maybe axially by temp change so it hit the target on a smaller part or whatever, amplitude of eddy probe reduced from  -4V to -3.4V and it
is very short so it didn't really work. Well my father refused to be part of the development of that monster so he maybe had a reason.
Normallly we have a tacho trig that track the peak at whatever amplitude it is but it refused to work on this.
It is a 30 year old build and a couple of decades older design or so, they didn't have so calibrated calculation models at that time......


__________________
Good Vibrations since early 1950's, first patented vibrometer 1956 in the US.
http://www.vtab.se
Alex

Avatar / Picture

Sr. Member
Registered:
Posts: 204
Reply with quote  #6 
Do you think absolute vibration Bode wouldn't work?
OLi

Sr. Member
Registered:
Posts: 1,915
Reply with quote  #7 
It is much combustion noise and blade pass but a filtered yes, we only got 1 chance. They must produce.
__________________
Good Vibrations since early 1950's, first patented vibrometer 1956 in the US.
http://www.vtab.se
Alex

Avatar / Picture

Sr. Member
Registered:
Posts: 204
Reply with quote  #8 
Also waterfall could be useful for analysis, no need for tacho signal.
OLi

Sr. Member
Registered:
Posts: 1,915
Reply with quote  #9 

I will look at what was collected, some FFT may exist.

I will tell them to monitor vector of 2xRPM and that FFT peak
that is about all I can do, a strange case I think.
There are no cracks right now I think but who knows when it
may take 10 years and the they found some other way to blow it up before that. 

So can a serious overload primarily torsion/torque create a significant stiffer shaft?
Funny it also became measurable longer that is several 1/10 of a mm on a few m length. 

It seems like the thing left is that shaft properties changed so crit speed moved up.
No bearing positions moved but all seals etc ended up in a very good position in this case
that may improve the situation and by the seal alignment influence the actual bearing "length"
and by that a shorter actual shaft length and by that increased crit speed and that may make most sense.
So the plan is to run, normally like 3 years to next overhaul unless anything else blows up,
fingers crossed. Have the car with hood in outward direction. 

 


__________________
Good Vibrations since early 1950's, first patented vibrometer 1956 in the US.
http://www.vtab.se
Previous Topic | Next Topic
Print
Reply

Quick Navigation:

Easily create a Forum Website with Website Toolbox.