More windy comments

I read, with some distress, the letters from the July 2007 issue of Structural Engineer. First, the letter from Patrick McGuire makes me nervous for the following reasons: 1) Yes, a separate analysis is required, just like for seismic versus wind calculations; 2) Yes, vertical loads are defined. The 10 pound per square foot (psf) is a minimum horizontal load listed in the wind chapter of ASCE 7. Thus, vertical loads conforming to Chapter 2 for strength and service load combinations listed are still required; and 3) Yes, steep roofs are penalized.

Most distressing is his reference to "nor is distribution among the building surfaces defined." The same method used to assign actual wind loads should be used to assign the 10 psf minimum to diaphragms, walls, etc.

I would be more prone to agree with his final comment on a more scientifically based criteria, but would be more interested in knowing, as a percentage, how many actual structures are designed for this 10 psf minimum. I would wager that it is fewer than one in 100, considering that no structure I have ever designed has been governed by the criteria.

Brian Johnson, P.E.
Via e-mail

Are computers replacing thinking?

I agree with many of the points made by James Parker, P.E., in his article "Where do we go from here? The evolution of structural design software" (which appeared in the August 2007 issue of Structural Engineer) regarding the use of software in the building design process. I especially agree with his warnings against improper use of software by inexperienced designers and lack of oversight by experienced designers.

As a plan checker for a building inspection department for Contra Costa County in California, I see plans and calculations by a lot of different engineers and architects. It is often clear that little time was spent analyzing the results of the reams of paper that spew out of the engineer’s computer. Many designers have a misguided sense of security seeing a computer printout giving beam sizes and shear wall types without a clear understanding of whether the results are reasonable.

I recently reviewed a set of computer-generated calculations that had the following conclusion at the bottom: "788 pounds > 789 pounds, OK." This article reminded me of another article that I read years ago in Civil Engineering titled, "The day the computer stole the engineer’s brain." We may still be living in that day.

Thomas Huggett, S.E.
Lafayette, Calif.

Hitting the mark

I really enjoyed "Is a ’spring in your step’ causing problems? Recommendations to minimize annoying wood-floor vibrations," by Frank Woeste, Ph.D., P.E., and Daniel Dolan, Ph.D., P.E., which appeared in the June 2007 issue of Structural Engineer. The article is the most concise and applicable summary I’ve seen on wood-floor vibration. I will confess that my usual approach has been to select floor joist based on a minimum delta of L/480 and pay special attention to solid bridging and gluing/screwing sub floors. However, when it comes to intermediate girders that are of moderate lengths, say up to 12 or 14 feet, I’ve been upping the delta to L/360, thinking that the chance of full loading—and subsequently uncomfortable vibration—is much less likely. I had not considered reducing delta to L/600—although I’ve done that on longer spans.

Woeste and Dolan’s article has opened my eyes to the situation, and I’ll be doing some "vibration analysis" on future floors. Thanks so much.

David R. Hourdequin, P.E.
Highlands, N.C.

Compensation comparison

Thanks for the information presented in the June 2007 article "If money grew on trees what else would you want from your job?: The 7th Annual Compensation Survey." It seems to me an important piece of information was missing in that nowhere was the average number of hours worked per week reported. It may be that an engineer in the Q1 category is working 40 hours per week. Another engineer in the median category might have to work 50 hours a week average, and may, therefore, be receiving a lower compensation when considering hours worked. This may also be reflected in the difference in earnings based on job title. A project manager may be working significantly more hours than a project engineer. It would be very helpful to include average hours worked with any compensation amounts.

Steven Call

Response from the editor

When we conduct this survey we ask respondents for their "current annual salary, not including any bonuses received." Additionally, when we do the analysis, we only include the data from respondents who indicate that they work full-time (as opposed to part-time or on contract) and at least 40 hours per week.

But to answer your question, the median number of hours worked per week from the 2007 survey is 40-45. Additionally, Q1 is 40-45 and Q3 is 46-50. We will include this data in future articles.