Using a Previous Design 2

[SKIAK]

A project I am working on is going to replace a fence. The fence is being replaced because, basically, it’s old, weathered, and falling apart. The fence is in a very remote location and it is very expensive to get materials and people out to the location, so the previous engineer used a unique type of anchoring for the fence; it was easy and cheap to transport and construct. I would like to use a similar (almost identical) design for the new fence, but I can’t actually calculate or verify that the existing anchor system works, but it seems to have worked on the previous design for at least 20 years. Nobody else that I have talked to knows how to calculate capacity for these anchors either (other structurals and geotechnicals).

But my question isn’t technical; I have talked with other engineers in the office and I have talked to geotechnical engineers that are involved in the project. I’m a relatively new engineer, just got my PE a year ago, and would probably be stamping this drawing. Different anchorage systems for the fence that have been suggested by other engineers and the geotechnical engineers would be much more expensive (on the order of 10x more from my cost estimates) for the sole reason that we can’t calculate or actually verify that the previous design “works.”

My immediate thought is that it seems to have worked for the past 20 years, and conceptually I can see how it works, I just can’t calculate it. It’s not a life safety issue, if the anchorage failed the fence would just fall over (I suppose there is a chance somebody could be walking beside it, but that is very unlikely) which of course is undesirable but I also think is unlikely.

Would it be unethical, or bad engineering practice, to use this previous design since I can’t actually verify it for myself?

 

[truckandbus]

This is a benign case considering the apparent low risk, but I'm concerned that as a relatively new engineer you might be on a slippery slope. "It hasn't failed before" is the same logic the folks at Morton Thiokol used to give the thumbs up for Challenger.

We all use codes and standards and cookbooks to do our work, but between education and experience they need to be applied correctly

 

[GregLocock]

A 20 year real life test across multiple units is a bit more than the O rings got. Incidentally the criterion Feynman said should have been used was to replace 'did not fail' by 'did not perform as expected', which would have dealt with the erosion issue.

So, since this is cost driven, is it worth inspecting all or some of the units in the field and checking that they have all performed as expected?




Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[SNORGY]

I think that the sketch provided by SKIAK would constitute a fairly good free body diagram that, at the very least, one could compute reaction forces from. Then the only question would be what soil conditions one would need to hold an eight-foot rebar in place in the postion(s) and angles shown.

Once that was done, one could go on to quantify a theoretical maximum wind speed, a safety factor, etc. until one obtained an "acceptability envelope".

If the results were acceptable, then certainly that would be good enough to stamp, would it not?

In the absence of calculations or proof of design acceptability by others, I would view it as incumbent upon myself to apply my own analysis, as right or wrong as that analysis might be, before considering stamping it. The other way to look at it would be, "OK, so if I have a fence that conceptually looks like this, how would I design it and back it up with calculations before I would be prepared to call it my own?".

The active word there is "own", because once you stamp it, you own it.

 

[truckandbus]

Not being critical of the design, just the decision to stamp the drawings...