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

Response to Keymark Message and additional Questions

[Subject Prev][Subject Next][Thread Prev][Thread Next]
Excellent questions since many of the list members have wondered about these
features. Let me ask one additional questions and then respond to some of
yours. I think it best to address them here rather than the Keymark list
only because most of us are at an evaluation stage and one persons question
can spur another to ask even more.

My question is that Model is suppose to allow for input of DXF files yet I
can find nothing in the help files or menus to explain how this is done?
Also, many features are listed in the help files as "Coming Soon". Can you
elaborate on these to some degree?

Plate Heights
I believe there is an assumption in the current code methodology that
regardless of elevation changes at any given level where the diaphragm is
stepped or offset - the entire footprint, except where a cutout or known
discontinuity occurs is monolithic in nature. If this is valid, the
transference of shear from one level to the next is assumed adequate. I
would think that the elevation change is minor (a foot or three) and that
the "pony wall" used to connect the levels has sufficient rigidity to
provide an insignificant difference in deflection. I'm sure where there are
conditions such as a Bell Tower, where this is not the case. However, for
the most part I think that the entire diaphragm is considered monolithic.
Therefore, the elevation changes would be disregarded and an average plate
height can be assumed for sloping or stepped roofs.
I think the manner in which Keylat treats this is valid as long as the user
has control of establishing the diaphragm height based on engineering
judgment or a conservative calculated average. Where the transition from
level to level is inadequate or is extremely flexible (such as the Bell
Tower example) there may need to be a creative way to address the extra
flexibility of the upper diaphragm - but this area will most likely still
act as a rigid element under the strict interpretation of the code.

Skewed walls:
Ask yourself this question - Is the location of the active fault and the
direction of the applied force always normal to the main axis of the
building? Of course not. It is only an assumption for the sake of the
analysis. Therefore, how can you discount skewed walls from having potential
value to resist lateral forces?  My judgment call is to assume that skewed
walls may be acted upon in either orthogonal direction. I can do one of two
things - design the wall as the sum of the force vectors OR design for both
directions and design the wall for the worst condition. I prefer the later
as it assumes the same premise that the rest of the structure assumes. If
the force is applied orthogonal to the skewed walls are all other walls in
the structure ineffective? I think not.

I consider skewed walls in flexible analysis as being orthogonal to the
direction of the load. Therefore, I calculate the reactions for each of the
two orthogonal directions and use the worst case result for the final design
of the wall. My reasoning is that the direction of the force is only assumed
and may be much different in reality. Therefore, there is no reason to
penalize the skewed wall by assuming the load to the wall as the hypotenuse
(the square root of the sum of the squares of each orthogonal load) of the
two vector loads.

In the rigid diaphragm analysis (as Dave Merrick pointed out to me) skewed
walls may pick up a negative component when both loads are applied
orthogonal at the same instant. One solution would be to ignore negative
components and apply only additive positive reactions.  Another way to
contend with this is to split the analysis into two parts - analyzing each
orthogonal direction separately and comparing results for the worst case
condition on the assumption that the force will occur in only one direction
at any given instant.

Dennis S. Wish PE