designing suitable stud to operate under vibration 5

[chumet1]

i would like assistance on how to design a governor valve positioner link/stud suitable to withstand the up & down system vibrations of 19.1mm/s as shown in the attachment. the positioner link stud shown in the attachment has fractured 3 times in the past year. when it fractured the second time we replaced the mild steel stud with a high tensile stud which snapped within 6days most probably due to its its brittle nature, it could not withstand the bending caused by the vibrations. we replaced the 6mm HT stud with an 8mm mild steel stud. i need assistance on how to design a stud to withstand the vibrations whilst awaiting the outage to investigate the root cause of vibrations. i would also want assistance on the vibration allowance limits on the governor valve.
THANK YOU...

 

[CoryPad]

Is the stud externally threaded over its entire length, with internal threads in both the actuator and link?

 

[MintJulep]

Move the bottom closer to one of the two big rods on the side and design a new top connection around the big rid to casting connection with a shorter cantilever.

 

[hydtools]

Clearly the stud is not sustantial enough for the loading. The base needs to be larger in order to handle the bending load. Perhaps a triangular or conical structure shape with a through stud.

Ted

 

[chumet1]

@corypad, yes that is exactly how the set-up is like
@mintjulep, i think u lost me there

 

[MintJulep]

Currently the little cylinder thing is positioned at the worst possible location.

Move it to the best possible location so that you don't need the long cantilever at the upper connection.

 

[CoryPad]

Currently the assembly has little to no pretension in the threaded rod, which means all of the vibration forces are transferred into the threaded rod which has a high stress concentration at the thread roots. Can you modify the actuator or link with a counterbore to allow for some rod stretch and pretension? The pretension will clamp the assembly together, and the friction forces generated will resist the vibration forces so that the rod does not fail by fatigue. I have attached a crude example. You will need a pretension method, either torque or linear.

 

[Dougt115]

The threads are a tremendous stress concentrator especially at the transition.

I would try a rod that is threaded only at the ends, use a torque nut on the rod to apply a preload and if possible have a counter sink machined into the larger structure (or weld on a fat sleeve) that the smaller part can recess into. This should remove or at least lesson the tranverse loading of the connecting stud.

 

[UchidaDS]

Not sure if you can use coupling to solve this problem.

 

[dhengr]

Chumet1:
Make a triangular steel plate stiffener/brace, which is coped to clear the head casting yoke. This should clamp over the large center nut or yoke casting, and it should run out almost to the tip of the positioning rod/link cantilever where it is clamped to the positioning link. This will settle down the vibration at the tip of the link. You might also clamp the link to the stiffener plate near its mid-length or back near the stud.

 

[Tmoose]

http://thumbs1.ebaystatic.com/d/l225/m/mbKfqd6tbalurSOYuyc4_mg.jpg

 

[byrdj]

I don't reconize the actuator, but I can assume this is a conversion from MHC to some digital controls. the 19mm/sec sould be some concern for a control valve, what freqency or displacement is this primarly at. Is it a noise problem in the controls?

the basic problem is typical mistake in linkage design. "Do not transfer force accross a thread". as suggested by several others above, you need to have a solid fit of the rod into the yoke. I would hope the yoke was spot faced so that the rod face was in full face contact. By boring this spot face deeper by the diameter of the rod with a slight interferance fit would be worth a try. I see a possible concern can become the thickness of the yoke and the depth of the thread does not allow for too much engagement before the actuator stem is contacted.

Is that a washer under the center nut?. If so, a better design might be to make an arm to replace the washer. this arm would need to have an indexing dowel into the yoke to prevent it from rotating when tightening the nut and hold alignment to the LVDT.

Another suggesting worth repeating would be move the LVDT to be mounted near one of the lift rods on the side, thus reducing the cantilever. the spring in the long cantilever can contribute to the control system noise and actually make the actuator hunting worse.

I would need to see more and possible a actuator/valve drawing, but it might even be simplier to move the LVDT off to the side and down lower beside the piston.

good luck

 

[byrdj]

the control valve modulation acceptance critera I used was less than 0.063 in/secs (1.6mm/sec). it was actually a curve of displacement versuses frequency. this was for a very robust design actuator/valve/sensor arrangement. this measurement had to be taken on the actuator and not from any control system signals. the maximum frequency I was looking for was 5 hertzs with most measurements done in periods. above 5 htz, the actuator is not fast enough to respond (but can still be damaged) and then one needed to look at pressures and servo demand currents to quantify the control system noise. the displacement at the actuator at 5htrz would be about 0.010"(0.25mm) for 0.063 in/secs

one of my wisdoms I state with problems related to steam turbine control instrumentaion is that;
1) the mounting of the attachment must be capable for a mechanic to use to pull himself up and stand on. &
2) the device must be robust enough that when removed, the mechanic can use it as a hammer.
(I don't think your design meets #1)

besides the modulation vibration, the design must be able to withstand the impact of tripping.

 

[tbuelna]

chumet1-

Attached is a sketch showing my suggestion for modifications to your link installation that could be incorporated fairly easily. Based on your description it seems like the current arrangement creates a large stress concentration right at the point where the stud threads enter the yoke. The dynamic vibration environment likely makes the problem worse by producing a high-cycle reverse bending load on the stud.

CoryPad's suggestion for a lengthened and preloaded stud is a great start. In my sketch I also suggest adding an enlarged base to the fixed end of the link, which will increase its moment capability. Tapering the link's cross section along its length should create a more even distribution of bending stresses.

Hope that helps.
Terry

 

[byrdj]

based on the additional information in the posting "boiler and pressure vessel engineering" concerning the vibration, as stated the source of the vibration needs to be identified along with its charectorestics.
It could be from the control system driving the actuator.
It could be resonance from the excessive turbine rotor vibration.
it could be resonance due to improper support, driven by acceptable turbine vibration
or it could be flow induce by the valve disk and seat.

for flow induced, is this a new turbine/design valve? if so, the OEM needs to get involved. If this valve has operated satesfactroy in the past (or on its sister unit) then there could be a problem with either the balance chamber, pilot valve or stem to bushing clearances

 

[CoryPad]

Great design Terry, that is what I had in mind with some nice extras.

 

[chumet1]

thank you all for the terrific ideas and suggestion. i really learnt a lot.

 

[byrdj]

would it be possible to tell the turbine make/vintage and if this was part of an upgrade or original?