About Roark's book: stress and strain

[kingnero]

Being European, I'd like to know whether it would be helpful getting me the "Roark's formulas for stress and strain" book.
Are the formulas in Metric or Imperial units?
No point in getting me a book full of inches, psi and lbs, as I do not come across these units when I work.

Anyone could confirm this?

Many thanks...

 

[EspElement]

i believe there is 2 versions of the book. Si and english

 

[PeterCharles]

But surely to an engineer it's principle that matters, you can always sort out the units according to what suits you.

 

[electricpete]

I have the 7th ed. It is advertised as being a mix of SI and "United States Customary Units":

http://www.bookmarki.com/Roark-s-Formulas-for-Stress-and-Strain-7th-Ed-p/007072542x.htm

From looking through it I see there seem many more USCU "examples" than SI examples, however all the formula's are presented in a standard engineering manner that is not unit-specific.

Re: PC's comment. FWIW I have lived my whole life in the US. I can/have worked most simple type problems in both systems. As the complexity of problem increases, I have found it is invariably easier to convert input quantities to SI (even though given in USCU), perform the entire calculation using SI units, and convert the output quantities back to USCU. By "easier", I mean fewer unit conversions requires, less thought and book-keeping required to tend to unit-related matters. Things like numerical differentiation and numerical integration flow naturally when SI units are used without extra consideration of factors related to units. My stuff usually doesn't include things involving empircal correlation relationships.


=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

I will also say the examples are a very small part of the book. I agree with PeterCharles from the standpoint that I personally wouldn't judge this book on the basis of the units used.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[IRstuff]

If you have Mathcad, you can get the Civil Engineering Library, which has Roark's formulas:

http://www.ptc.com/products/mathcad/civil-engineering-library

While they're entered with US units, Mathcad does unit conversions on the fly, so you simply enter the correct values and SI units, and the equations will still work correctly.

TTFN

FAQ731-376

 

[kingnero]

I might be persuaded and order me a copy, still not very convinced as graphs with psi in the y-axis doesn't tell me much... but indeed it's the principle that matters.

 

[IRstuff]

Conversion of units is trivial in Mathcad, if that's what's bothering you.

TTFN

FAQ731-376

 

[IRstuff]

Even femtograms per square furlong, if that floats your boat.

TTFN

FAQ731-376

 

[kingnero]

Don't have mathcad, I do about all of my calculations by hand using an old casio calculator...

 

[zdas04]

I happen to be working through some of the Pressure Vessel calculations in Roark's in a MathCad e-Book this morning, and I surely would hate to have to go back to solving this stuff with a slide rule (or even a calculator). Stress on a flat plate (which I recalled as two equations) is two pages.

David

 

[IRstuff]

For a limited set of calculations, one can easily imagine programming a calculator or entering the problem into Excel. But, after about 6 months or a year, someone comes in with requirements in a different unit system; then what? In most other cases, you wind up effectively recoding the equation, just to make sure you got all the conversions correct.

In Mathcad, that's completely unnecessary:
> The equations are in natural form, so they're completely readable as recognizable equations like those in the text
> In most cases, an entry requiring, say, psi, can simply be entered with Pa, and that's ALL you have to do.

Obviously, much of this requires that the equations be properly entered and dimensioned in the first place, but the effort is not that different than validating the equation as used in any other calculation approach. The difference is that if done correctly, you ought not need to do that ever again.

One option, to get your feet wet, is to buy a copy of Studyworks for about $10:

http://cgi.ebay.com/StudyWorks-Math...emQQptZLH_DefaultDomain_0?hash=item3efd66da83

While it's quite outdated, ca. 2002, this stripped-down version of Mathcad is still quite useful for numerical calculations, and provides a human-readable input/output for archival and documentation purposes. Note that Studyworks does not include programming and some higher-order math functionality, and only a limited symbolic capability.

TTFN

FAQ731-376

 

[JStephen]

Get the book, it's useful either way. Most of the formulas are not unit-specific; right offhand, I can't think of any that are.

 

[Ussuri]

I have the 6th International Edition. Most of the formulae are not unit specific. The examples are a mix of both SI and Imperial units.

The preface even states "The sixth edition uses the International System of Units (SI) in presenting many of the example problems. Tabulated coefficients were in dimensionless form before and remain unchanged. Design formulas drawn from works published in the past remain in the system of units originally published or quoted."

Its a useful reference so you should get a copy.

 

[rb1957]

i'll be contray ... if you don't want to convert units, then get a book with only SI units ... just be careful of Newtons and kgf.

 

[IDS]

just be careful of Newtons and kgf.

A kgf is not an SI unit!

(Neither for that matter is the cm.)

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[zdas04]

Isn't that hilarious? The first time I saw a gauge marked in "kg/cm^2" I took a picture of it to show people when they start on the "Metric is Perfect" high horse.

I work with metric a lot. I work with US Traditional Units a lot. Jumping back and forth is pretty easy (especially since I started imbedding the required units within empirical equations in MathCAD--I love that I can input "mm" and then MathCAD divides it by "inches" to get a unitless number that is consistent with the constants).

I think that the important thing with units is that you have clear communication. If you do that then microFarads/fortnight is almost reasonable.

David

 

[CoryPad]

IDS,

You are correct about kgs, but incorrect about cm.

http://www.bipm.org/utils/common/pdf/si_brochure_8_en.pdf

http://physics.nist.gov/Pubs/SP811/contents.html

 

[IRstuff]

Dave,

But shouldn't it be furlongs/fortnight? Both were units added in Mathcad version14.

TTFN

FAQ731-376

 

[electricpete]

To zdas - I’m not sure the exact definition of “metric”, but I can guarantee cm is not part of the SI system.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.