Discharge ring cacks

[engdrve]

Looking for advice
I am a mechanical engineer that works in a hydroelectric plant with Kaplan turbines of 180 MW. We have a problem of appearance of cracks in the discharge ring exterior horizontal reinforcement’s plates. Transverse cracks are also appearing in the bimetallic welding between the discharge ring and the cone of the draft tube, these cracks are located right under the joints of the discharge ring sections. The discharge ring has an inspection gallery for the whole surroundings, which means that only an eighth of the discharge ring is embedded in concrete, the diameter of the discharge ring is 7.8 meters. We have measured the vibrations of the ring and the maker has informed us that they are inside an acceptable range, between 150 and 250 microns during stabilization, and between 50 and 100 microns when the power has been stabilized. Frequencies have been measured and none considerable differences has been found. I will appreciate any advice from engineers that have experienced this type of problems before or have knowledge on this type of problems. What could be the cause of the generation of cracks and which could be the possible solutions?

 

[bklauba]

Do you have details about the discharge materials and manufacturing processes?. Please post materials (bi-metallic, is this a fab made of lo-C and 300 series SS?), whether PWHT or other form of SR was used, etc.

Also, what does the builder of the equipment (that gave their opinion on the vibe levels) say?

BK

 

[engdrve]

The discharge ring material is ASMT A240 and the discharge reinforcement’s material is ASMT A537. Bimetallic is a site weld between ASMT A240 (Discharge) and ASMT A36 (draft Tube). The discharge is fab made with PWHT (Temp. 605,00 C , holding 3-4 h) . The first instruction from the builder was to grind the mouse steps a make it round since stress concentration was occurring there and causing the cracks. It did solve the problem of cracks on the mouse steps but the cracks started to appear on the edges of the horizontal plate reinforcement and on the above mentioned bimetallic weld. So the builder will continue on with his investigation. I understand from you it is difficult to come out with a solution with so little data, but I will appreciate any clues to continue on with our investigation.

 

[metengr]

engdrve;
The material specification ASTM A 240 is for a general group of stainless steels. As I was reading through your post, the bimetallic you mention is considered a dissimilar metal weld. What you need to do for starters is to have the contractor provide you with a copy of the weld procedure(s) that was used to join the dissimilar materials. Review the selection of filler (weld) metal. It should have been 309 stainless steel.

 

[engdrve]

yes, The filler (weld) metal is E309. I have the weld procedure, and i think it has been done properly. I think the cracks on the in the bimetallic weld are small and not the main issued. Because the cracks on the carbon steel reinforcement plates Themselves are worse. So I think this is a matter of structural design. I am looking for somebody who could have experienced something similar ia a kaplan turbine.

 

[metengr]

engdrve;
Ok, great one less headache to worry about. One other concern to check, are the cracks in the steel reinforcement plates running transverse to the weld or parallel to it? If the cracks are confined to the weld and extend into the base metal and are transverse to the weld, you may have had a fabrication problem with hydrogen cracking. Just something to confirm.

 

[engdrve]

metengr, Thank you for being interested in the topic.
Yes the cracks in some cases extend into the base metal.
forgive my ignorance but what is a hydrogen crack. I'm beginning to believe that, we are dealing with a premature problem of material fatigue due to the discharge vibrations. Another thing, I have not mentioned that the concrete around the section of the discharge ring that is embedded is being demolished little by little and the ring seems to be getting loose. The machine is still under guarantee. we are waiting for a solution from the builder, but we want to be prepared because the solutions that the builder has been giving us have not worked. if you want any other information let me know. thanks again.

 

[metengr]

engdrve;
Hydrogen cracking occurs either during welding or shortly after welding (this is called delayed cracking or hydrogen embrittlement). This damage mechanism occurs when atomic (nascent) hydrogen generated from the interaction of the electric arc (from welding) and moisture in the air or adjacent to the steel surface enters the weld metal. Nascent hydrogen will diffuse to regions of high residual tensile stress and result in cracking, as a means of relieving residual tensile stresses. Causes of delayed cracking are; lack of preheat control, poor cleaning techniques and poor weld rod maintenance.

Typically, hydrogen embrittlement cracks are oriented transverse to the weld and are either confined to the weld region or continue into the steel base metal.

We have a problem of appearance of cracks in the discharge ring exterior horizontal reinforcement’s plates. Transverse cracks are also appearing in the bimetallic welding between the discharge ring and the cone of the draft tube, these cracks are located right under the joints of the discharge ring sections.

I most certainly agree with you could be having fatigue crack propagation issues related to poor design. However, I wanted to alert you to other damage mechanisms to consider before you embark on re-design (especially with a warranty issue). I have been down this path before.

 

[engdrve]

metengr,
thakns for the alert. I see you understand what is a warranty issue. That is why we have to check every thing. Sorry that english is not my firts language and i have hard time explaining the problem and I could be misssing information. Seems like the next proposal from the builder will be to use the gallery that surrounds the discharge ring in order to set jacks between the ring and the gallery´s concrete wall, to restrict the movement of the ring. I will be scare of the problems being transfered to the concrete. And then the jacks breaking the concrete and getting loose againg. well, any commet I will appreciate it.

 

[EdStainless]

If you have enough motion that you are loosening the ring I suspect that it is not stiff enough. If it is flexing too much that could drive crack propigation. It is not uncommon for small cracks to be found in these structures, but a solid ring with concrete backing should not be moving.

What could be different compared to your other units:
Either this ring had too many imperfections (or they were too large.
This ring may not have been supported properly by the concrete.
There may be a slight difference in flow causing a flow related vibration in this unit. (this my not show up as turbine vibration)

I agree, I don't like the jack solution. If there is room (or they can cut room) to insert jacks then why not pressure grout the ring in place? Using jacks will cause higher local stress on both the ring and support concrete, heither of which is desirable.

= = = = = = = = = = = = = = = = = = = =
Rust never sleeps
Neither should your protection

http://www.trent-tube.com/contact/Tech_Assist.cfm

 

[engdrve]

EdStainless,
All ten Units are beginning to present the same problem.
We can not pressure grout the ring completely because it is necessary to leave an access gallery around the ring in order to be able to reache under the wicket gates. since on this design the wicket gates´s gaps are adjusted from under. Right now only an eighth of the ring is embedded. We have now the uncertainty if the ring is too stiff and needs to be loose or the ring is to flexible and needs to be stiffer in order to avoid the cracks.

 

[lsawyer]

I think that the implantation of a shock-absorbing mechanism is the solution, it would not change the enough of the ring, but it diminishes the displacements since the energy dissipates through this mechanism; like happens in the car, when a rim (wheel)hits with a hole.

the fatigue is an unpredictable phenomenon, and a mechanism dissipator of energy to resolve this problem. I wait for its commentaries, thanks.

 

[bklauba]

Isawyer: What is the (important?) word omitted, I think, from your reply, before the word "enough".

Is your suggetion a practical solution for such a device? Also, does this help the OP; i.e., can such a solution be implemented without a major deconstruction/reconstruction of the facility? Keep in mind that the vibration levels appear (at least at last word from the OP) to be acceptable, at to the hydroturbine manufacturer.

I was tempted to suggest things that could have been done during the manufacture of the hydroturbine, but such suggestions do not help the OP; it is too late for him to implement such ideas.

BK

 

[lsawyer]

i am sorry, the English is not my native language, and frequently I confuse some terms.

"enough" is not the right terms. stiff is the right term.


about your coment, I think that this problem needs the implantation a new mechanism. a turbine of 7.8 mtrs of diameter is very great, and the loads on the ring also are it. but you are right. it is too late for him to implement such ideas.

what is your proposal?

 

[bklauba]

I think a good number of people have looked at this thread and not been able to make suggestions that address the problem. My "solution", but too late to implement now, would have concentrated on making sure that the rings were dimensionally stable prior to installation, perhaps (if appropriate) along these lines:

http://www.vsrtechnology.net/prelaunch/images/stories/VSRpdfs/Voith_Siemens_2.pdf

. . . . but once installed, good "in situ" repair work is likely the only means of addressing the problem, but is hardly a real solution. I will allow others who participate in this forum who have much more to offer on the subject of weld repair and related topics to offer proposals.

BK

 

[lsawyer]

very interesting. I think that the study of the ring must continue. as it is not possible to disassemble the ring, it is necessary to implant a mechanism, my point is, the repairs are not going to give solution, it is necessary to change the conditions of the ring, or to change the design. however, i really like listen all the proposals of the people of the people.

 

[bklauba]

But the OP stated that the turbine builder had measured the vibration levels, and that they were found to be acceptable. It is hard to see how a "shock-absorbing" mechanism, coupling, etc., would have solved the problem, since there is no evidence of excessive vibration levels.

My fear is that the rings were dimensionally unstable (some machinists use the term "live" to describe this), and that during the course of installation and usage, the rings stress relieved themselves. In the course of doing so, they distorted, or, constrained by the concrete and mating components, developed preloads. This is one explanation for the cracking seen by the OP in the reinforcement plates. Are there other explanations? Of course, and these have likely been examined (material, welding procedures, etc.), but the OP offered no finding that these were possible culprits. Rings of this type, if not effectively stress relieved, will distort, either during transport or after installation. The quandry now is, what to do now?

The question remains: Are there other likely explanations for the cracking?

BK

 

[EdStainless]

Well I still go back to the fact that some portions of the ring are very stiff and well anchored, while other sections can flex a lot. I would tend to believe that the vast differences in local stiffness lead to some overloading.

If the builder wants to install jacks I say let him. But make sure that they also install a system to detect further failures. You should have a remote system to assure you that everything is still tight.

What will they do when a ring cracks clear through?

= = = = = = = = = = = = = = = = = = = =
Rust never sleeps
Neither should your protection

http://www.trent-tube.com/contact/Tech_Assist.cfm