Transmission tower reinforcement 2

[towerengr]

I am reviewing a design submittal for reinforcing a high voltage transmission tower which is in service. All structural members are single angles, mostly equal leg.The existing member is overstressed in compression and the proposed reinforcement consists of adding a member of the same size by bolting it to the existing member before the existing end connections. Between the end connections a single bolt connection (connected to one leg) is used at 3 or 4 locations. There are no intermediate lateral braces and I am concerned that the load transfer between the existing member and new reinforcement may not take place. Also the new member is braced by the existing member only as it is not connected to the end supports.

The bolt holes are oversized so the new member can slip with respect to the existing member.

Any recommendations on how to establish the connection requirements and the effectivenss of this reinforcement will be much appreciated.

Member lengths vary - may be from 6 ft to 30 ft. Bolt holes are 11/16 or 13/16 inches in dia with 5/8 and 3/4 inch bolts respectively.

The applicable code is ASCE 10 but this issue is not addressed. I do believe that Newmark's iterative solution may apply here but if I can figure out how the load distribution works it will be nice. The other possibility is the l/r is reduced because of the new member and the load stays in existing member.

Sorry about the long post.

 

[hokie66]

It would help to know the proposed orientation of the reinforcing angle to the existing angle leg.

 

[towerengr]

Hokie66,

Thanks for a very quick response.

Let us assume the existing member angle as having horizontal and vertical legs. It is connected to the tower framing with the horizontal leg. The reinforcement member is connected to the vertical leg.

Hope this explanation is sufficient. Otherwise, I will scan the drawing and forward a sketch.

 

[hokie66]

A sketch would help. I had assumed it was the tower legs which were in trouble, but on rereading it seems it is bracing members. Correct? When adding an angle to another angle for strengthening, it is generally best to make the combined member into a "zee" shape.

 

[dik]

How much are they deficient? Can the remedy be accomplished by welding a BAR/Plate across the toes of the angle. This will improve the r value as well as the area.

Dik

 

[towerengr]

I apologize for the long delay in getting back. A scan of the reinforcement detail is attached. The overstress is around 25%.

My question to the forum is about the underlying assumptions with this type of reinforcement. Can we add a member to an existing member as shown on the attachment and expect it to reinforce the member to prevent buckling? What analytical technique/method should one use to compute the allowable load?

Thanks.

Sam

 

[sdz]

If you are trying to make a composite member you want to eliminate slip in the joint. Consider using something like Huck bolts. Even then the centroid of the composite section will be different to the single angle and the effect of that should be checked. Are 3/4" bolts getting a bit big for 2" angles?

 

[towerengr]

Sdz,

Thanks for your response. This is the answer I was looking for.

This is my position too but I am having hard time convincing the young and old(?) engineers that slip will prevent the composite action. They say not to worry!!

I prefer the new member be connected to both end joints so that the new member loading is never in doubt.

 

[hokie66]

I would tend to agree with the other engineers in your office. Just tighten the bolts, and the new angle should act as a brace for the existing. Take the stiffening angle in as close as practical to the joint, and bolt them together at close enough centres so the combined section controls.

I didn't see a bolt size on your sketch, but agree that 3/4" is too big.

The holes in the existing would have to be drilled, not punched, which will be difficult with these very small angles. You should talk to an erector as to the feasibility of the installation.

 

[transmissiontowers]

The 3/4 bolts in a 2x2 is done all the time. We use A394 bolts which have smaller heads than the A325 than most people are used to. We also have portable hydraulic punches to do the holes in place.

This looks like a lacing member that could be replaced with a larger size instead of doing the bolt on. I put up PCS sites on our towers and have to analyze them for the added loading and we replace lacing members all the time. I try to stay with the same size on the bolted leg and increase the OSL. In your case I would try a 3x2x3/8 and see if the compression allowable comes up enough. If there is enough room, you can go to a L2.5x2.5 to increase the L/r.

Do you use PLS-Cadd's Tower program for the analysis? If so, there is a interactive member size function that I use all the time but don't let it select more bolts.

If you must use the added angle and the original 2x2 is OK for the tension load and the bolt shear and bolt bearing, then all you need to do is increase the radius of gyration to lower the L/r. Since it is not a P/A problem you don't need to transfer any compression load into the added splint angle. The splint is only there to increase the Euler Buckling allowable. It does need to be stitch bolted so it is effective. ASCE 10 section 3.11 talks about stitch bolts.

_____________________________________
I have been called "A storehouse of worthless information" many times.

 

[hokie66]

transmissiontowers,

Glad to hear from someone who has done this, and thanks for the correction about my comments concerning the bolt size and punching in place.

Replacing the lacing members sounds like a much simpler solution to me.

By adding the same size angle in the manner indicated by the OP, the radius of gyration in one direction is not changed. Making the combined member into a "zee" section as I suggested above would increase r.

 

[transmissiontowers]

Yes, you have to recalculate the L/r in all 3 axis to see which one controlled for the built up section. The z-z axis (weak) usually controls the compression but not always. Any brace point is a zz brace point and the crossing diagonal point can be assumed to be an out of plane brace point if the tension member has 20% of the load value in the compression member you are bracing. Take a look at ASCE 10 Example 7, 8 and 9 page 58 for some info on crossing diagonals. The general idea is the tension member can provide some out of plane resistance if it has enough tension in it, but if both diagonals are in compression (very rare) for a particular load case, the crossing point cannot provide a brace point. PLS-Tower has an automatic check of this and can adjust the bracing accordingly.

If you really want to get picky, the Euler curves and equations are based on a constant cross section and having only the middle 3/4 or 7/8 with a higher r might be a problem. IIRC, the curves are based on testing lots of single angles in compression and I'm sure they didn't test partially built up members. This is why I suggested replacing the member with a larger one.

If you are worried about removing a member in a tower with wires on it, we do it all the time (not legs, just diagonal lacing). The tower is controlled by a high wind or an ice load and if you switch out one at a time with a relatively calm wind, the tower will be fine. There are multiple load paths with the missing members out and the load will just re-distribute somewhere else. If you are really worried, you can remove the member from the model and run a load case with a 20 MPH wind and see what the stresses are in the remaining members.

In summary, I would see if replacing a member will solve the problem first. If you can't get one to fit and must use the extra angle in a Zee or Tee geometry, I would use the built up section L/r and make sure the new allowable compression load was about 1.25 times the compression load in the member using the Euler equations. I have no basis for the 1.25 other than a gut feeling that it should be stronger by that amount.

If the client insists on proof, you can do a full scale test of the tower with the revised member installed and test to destruction, but this is pretty expensive.

_____________________________________
I have been called "A storehouse of worthless information" many times.

 

[towerengr]

Hi Transmission towers,

I have been at this long enough that I know a thing or two. I am aware of everything you wrote about and left to me, I prefer a new member also. But the client would like to leave the designs alone unless one can prove that it will not function as assumed. That is why I was seeking analytical techniques to review the proposed addition of a new reinforcing member that is not connected to the supports. As you say, the ASCE 10 does not address anything like this and it is a grey area. My posting in this forum was aimed at getting some response from experienced engineers who might have tackled a similar problem.

Full scale load test is from the old days. Nowadays everybody relies on the computer programs. During the ASCE conference on Transmission structures in Dallas last fall, there was a paper that talked about full scale load tests on family of towers conducted by a utility and towers experiencing member failures after the structures were designed and checked by a commerically available software.

Your input is not worthless information. If you are in the tower business, it is awful lonely and when a tower question comes along, we cannot just answer the question!

Thanks

 

[transmissiontowers]

I was not sure about your level of experience with towers and I usually see questions about towers in this forum:

http://www.eng-tips.com/threadminder.cfm?pid=608

I've been analyzing and designing towers since 1973 for a utility on the gulf coast and I was at the ASCE conference in Ft Worth last year, so we may have bumped into each other there. I know most of the heavy hitters in the tower business, but you are correct that there are just not many tower engineers out there. I also do substation structures and was vice chair of the recently published ASCE 113 on substation structure design. ASCE 10 is in final balloting for the next revision and should be out this year or next.

Getting back to your question, I think you would be safe in stitching another angle onto the original. If you use PLS-Tower, I would use the same properties as the single angle for area, thickness, holes out, number of bolts, etc but put in new values for rx, ry, rz based on the combined section. I would probably create a new group in the xxx.ANG part file and call it SAS and put in the properties you calculate. I would still shoot for a minimum %used value of 80% to make sure it is strong enough. The extra angle is just there to increase the compression allowable by increasing the rx, ry, rz.

You might also just switch the original member type to Tension-Only to see how the loads redistribute if the original L/r was close to 250 or 300. The T-O designation will limit the compression force in the member to 100% of its allowable and shed the extra load to the other members in a geometric non-linear analysis.

_____________________________________
I have been called "A storehouse of worthless information" many times.

 

[towerengr]

Transmissiontower,

Thanks for leading me to the right forum to post tower related questions. I concede that you are ahead of me, but by two years only!!

I do not know whether you worked with the BPA tower analysis program. I enjoyed the challenge of working with a program where to get a plot you had to wait 6 hours and there were no beam member groups. But it had one advantage: you could analyze body panels and extensions in one run!!

I guess my input is also useless for most of the forum!

 

[transmissiontowers]

I tried ATADS once upon a time and I know Leon (chair of ASCE 113) very well. If you can ever get a hold of him, he might know of some paper about your problem. His PHD research was in transmission towers and he is one smart guy.

My useless info tag was given to me by co-workers who know I am a trivia nut and have seen many TV programs and movies, and at one point I used to know the firing order of Chevy V8 engines.

_____________________________________
I have been called "A storehouse of worthless information" many times.

 

[towerengr]

Transmission Tower,

You then must be Sherlock Holme's good friend!! I know Leon when I was in my previous job. Of course, he is smart. I have kept a low profile and did not bother to chase after committee memberships.

In my book, Massimo Finzi, Chief Engineer SAE (70's and early 80's) was one of the sharpest tower designer/engineer I met. I also have lot of respect for the maverick TLine engineer from Montreal who designed skyhooks!!

This post is way off subject but I will keep trying to find an answer to my problem through numerical methods. It is lot of hard work trying to read old text books!

 

[transmissiontowers]

If you are into FEM I may have the solution for you. We use GTStrudl and in the advanced analysis course I took, there was a demo problem where a single angle was modeled up with plate elements for the entire length. You apply a small axial load and do a buckling analysis in GTStrudl. The output is a buckling factor which you multiply the input axial by to get the allowable buckling load. So, you could model your composite stitched angle connected at the bolts, apply the axial load to the heel of the original angle and run a buckling analysis to find out the allowable compression load. Of course run the original angle to see how it compares to the calculated allowable you get from ASCE 10.

You might also contact Mike Miller who used to be at BPA and is now at SAE Towers in Houston with Klaus Sutter (sp) who has designed many towers and may know about stitch angles. Call Leon and see if he knows about any papers.

_____________________________________
I have been called "A storehouse of worthless information" many times.