Need a book? Engineering books recommendations...

Return to index: [Subject] [Thread] [Date] [Author]

RE: DeVere (vs. Madden)

[Subject Prev][Subject Next][Thread Prev][Thread Next]

Again, I received an undeliverable error on my messages to you, yesterday,
december 2nd..So we will have to postpone our personal chats until we can
communicate again. Let me know if that alternate email works.

About your foundation beams. . .

The book I have, "Foundation Analysis & Design" by Joseph Bowles calls the
Winkler Method the "Classical Solution of Beam on Elastic Foundation" It
gives no credit to the developer of the Finite Element Method other than

Here are the disadvantages listed for the Winkler/Classical Method vs. the

1.Assumes weightless beam (but weight will be a factor when footing tends to
separate from the soil)
2. Dificult to remove soil effect when footing tends to separate from soil
3. Difficult to account for boundary conditions of known rotation or
deflection at selected points.
4. Difficult to apply multiple types of loads to a footing.
5. Difficult to change footing properties of I,D, and B along member
6. Difficult to allow for change in subgrade reaction along footing. 

The author also states that "Rarely is the classical approach more worthwile
than the FEM"

Let me know when you get a new email, I've given up after 20 tries to

Talk to you soon (hopefully!)

Gerard Madden, P.E.
Civil Engineer, Associate
CRJ Associates, Inc.
email: gerardm(--nospam--at)
tel: 650.324.0691
fax: 650.324.0927

> -----Original Message-----
> From:	Bogdan [SMTP:prodomus(--nospam--at)]
> Sent:	Friday, December 03, 1999 6:08 AM
> To:	seaint(--nospam--at)
> Subject:	Re: DeVere (vs. Madden)
> Dear Mr. DeVere,
> I would appreciate very much if, in answer to my message, you could tell
> me
> the principles (theory) on which american calculations regarding
> foundation
> beams are based ; I doubt they are entirely "strange" of "these authors"
> I've mentioned.
>     However...:
>         JEMOCIKIN is the (russian) author of this calculation method based
> on the hypothesis of "elastic semi-infinite space (or plane)", being
> considered only vertical forces on beam (plate); this calculation method
> involves the sharing of beam in a finite number of "panels", every one
> being
> supported by a spring; problems: does this model (very much like a fem
> model) describes correct the curves of displacement? Is it possible to
> have
> a correct definition (value) for spring's constant ?
>         FUSS-WINKLER is a method developed considering a continuous link
> between beam and ground, involving the equivalence of ground and beam
> displacement for every point; it is established the formula p=Ks*y, where
> Ks
> is the rigidity coefficient for the ground and "y" is ground defformation;
> practical calculation is based on the differential (4-th degree) equation
> of
> an infinite beam deformed axis on ground. In order to transform the
> infinite
> beam to a finite one (usually having both ends free), the beam is
> elongated
> at both sides with virtual pi/4 distances, on wich are applied known
> virtual
> forces choosen in such manner to respect end conditions imposed by real
> problem (usually null end bending moments and shear forces). All
> calculations are very accurate IF Ks is well defined (i.e., for uniform
> granulation sand Ks=5.52...13.8, meaning more than double !!!)
>         UMANSKI-KRILOV method is also known as "initial parameters method"
> being recommended for "finite length" beams (alpha*L<5); it is also based
> on
> 4-th degree differential equation of an infinite beam, where for every
> constant of integration is given a phisycal meaning; the results are also
> very accurate, but more easy to obtain by manual calculations, who can be
> made for any desired section (the method can be assimilated to "Ritter
> section" method for girder poles).
>     Even if both UK and FW methhods use the same initial equation, the
> results are (sometimes) very different; I have found UK being the most
> satisfying method because its simplicity and range of applicability.
>     But WHY these differences for some ACCEPTED calculation methods? Where
> is the calculation accuracy reflected in solutions? In these condiyions, I
> could simply meka a choice for some no matter what dimensions, looking for
> the appropriate method of calculation AFTER !
>     The same thing can be done in case of using differential or FEmethod
> (due to necessity of defining of Ks).
>     Do you know any method to define it correct?
>     Thanks in advance.
> Bogdan Dumitrascu - no title
> Mr De Vere replied to a personal message of Mr.Madden:
> >
> > I have been able to find only brief citations of these authors in the
> > foot notes of my reference books. Books by the authors seem to have been
> > published in Russia circa 1935.  Would you be willing to share this
> > information with us?   Your efforts will be greatly appreciated.
> >
> >
> > Best Regards,
> >
> > Peter  De Vere
> > devere(--nospam--at)