PSN-Friends--I have read with interest the exchanges regarding stability =
and modifications to the so-called "Lehman" seismic design.  The plans =
as published in July, 1979 Sci. American were the result of 5 years of =
running the system in a number of environments, ---if I remember we =
tried 4 different types of Milne cantilever long period units, and =
settled on the design published.  As I have said before, the system--as =
most seismic designs--lends itself to modifications & improvements.
   The BallBearing and Tensional Wire designs for the boom end sound =
ingenious--although I have never tried them.  I stayed with the 15-cent =
hardware 5/16 or so bolt ground to a knife edge.  When mounted properly =
against a hard plate or flattened bolt head epoxied to the upright, I =
never experienced any undo friction or flattening of the knife edge.  =
Microseisms were always there--in fact one would have to run fast to get =
away from microseisms.  One amateur enthusiast in Puerto Rico had to =
settle for his system on the 14th floor of a high-rise apartment =
building.  In windy weather the building swayed, and the microseisms =
just "rode the swells" as he put it.
    It is easy to "overkill" the location of the sensor.  Ground level =
on a concrete floor with little temperature or humidity changes is =
ideal.  Direct connection with bedrock is not necessary.  One of the =
best locations during the testing time was in an unused room of a =
laboratory building.  Several feet of fill clay had been  leveled, and =
several inches of gravel on top.  Any bedrock (limestone here is =
Virginia) was probably 10 or so feet away.  We laid down 3 small garden- =
shop masonry step-stones------mounted our system--let it settle for a =
day or two, and began recording.  Later the area was finished with =
concrete floor, etc, and placing our sensor there gave similar =
satisfaction--only now more building noise was evident--vibrations from =
the Chiller 150 ft. away, and stresses in the building and stairwells as =
students moved to and from classes.
   I do feel a kinship with persons frustrated in setting up and =
stabilizing a system especially in a closed space.  The Sci. American =
design should be stable for weeks at at least 15 second period.  A good =
heavy base material other than wood is best.  Composite material like =
laboratory table tops is ideal.  Rather than have adjustable tri --feet =
on the base, I would suggest firm solid feet--fixed bolts or similar, =
and then use thin shims to complete the task once you have the mechanics =
of the system in the "ball park".  Five or ten mil thick sheet metal =
pieces work great here.  Once you have a 10 second or so period =
centering ok, then all you have to do is add shims to the "front" leg =
and the period goes up in a nicely until you reach instability--then =
back off a shim or two.
    If there is a steady trend to drift to one side, chances are the =
base is tilting a bit  due to a structural weakness---or more likely the =
slab on which the sensor is placed is moving.  I have known of systems =
placed in the corner of a home basement, and periodically drift was =
noted as the house foundation settles a bit--and this can happen over =
years and not be detected  by sight.  If you really wish to overkill =
your sensor base you can do what amateur astronomers do--pour a concrete =
slab block to attach their scope base, and surround the block with =
several inches of sand---then there is no walkup tilt of the base, and  =
any lateral mechanical vibrations at minimized as well.=20
   Well I have rambled enough--good stability to all, and Season's =
Greetings!

                                                                      =
Jim Lehman=20







 
PSN-Friends--I have =
read with=20
interest the exchanges regarding stability and modifications to the =
so-called=20
"Lehman" seismic design.  The plans as published in July, 1979 Sci. =

American were the result of 5 years of running the system in a number of =

environments, ---if I remember we tried 4 different types of Milne =
cantilever=20
long period units, and settled on the design published.  As I have =
said=20
before, the system--as most seismic designs--lends itself to =
modifications &=20
improvements.
   The =
BallBearing and=20
Tensional Wire designs for the boom end sound ingenious--although I have =
never=20
tried them.  I stayed with the 15-cent hardware 5/16 or so bolt =
ground to a=20
knife edge.  When mounted properly against a hard plate or =
flattened bolt=20
head epoxied to the upright, I never experienced any undo friction or =
flattening=20
of the knife edge.  Microseisms were always there--in fact one =
would have=20
to run fast to get away from microseisms.  One amateur enthusiast =
in Puerto=20
Rico had to settle for his system on the 14th floor of a high-rise =
apartment=20
building.  In windy weather the building swayed, and the =
microseisms just=20
"rode the swells" as he put it.
    It =
is easy to=20
"overkill" the location of the sensor.  Ground level on a concrete =
floor=20
with little temperature or humidity changes is ideal.  Direct =
connection=20
with bedrock is not necessary.  One of the best locations during =
the=20
testing time was in an unused room of a laboratory building.  =
Several feet=20
of fill clay had been  leveled, and several inches of =
gravel on=20
top.  Any bedrock (limestone here is Virginia) was probably 10 or =
so feet=20
away.  We laid down 3 small garden- shop masonry =
step-stones------mounted=20
our system--let it settle for a day or two, and began recording.  =
Later the=20
area was finished with concrete floor, etc, and placing our sensor there =
gave=20
similar satisfaction--only now more building noise was =
evident--vibrations from=20
the Chiller 150 ft. away, and stresses in the building and stairwells as =

students moved to and from classes.
   I do =
feel a kinship=20
with persons frustrated in setting up and stabilizing a system =
especially in a=20
closed space.  The Sci. American design should be stable for weeks =
at at=20
least 15 second period.  A good heavy base material other than =
wood is=20
best.  Composite material like laboratory table tops is =
ideal.  Rather=20
than have adjustable tri --feet on the base, I would suggest firm solid=20
feet--fixed bolts or similar, and then use thin shims to complete the =
task once=20
you have the mechanics of the system in the "ball park".  Five or =
ten mil=20
thick sheet metal pieces work great here.  Once you have a 10 =
second or so=20
period centering ok, then all you have to do is add shims to the "front" =
leg and=20
the period goes up in a nicely until you reach instability--then back =
off a shim=20
or two.
    If =
there is a=20
steady trend to drift to one side, chances are the base is tilting a =
bit =20
due to a structural weakness---or more likely the slab on which the =
sensor is=20
placed is moving.  I have known of systems placed in the corner of =
a home=20
basement, and periodically drift was noted as the house foundation =
settles a=20
bit--and this can happen over years and not be detected  by =
sight.  If=20
you really wish to overkill your sensor base you can do what amateur =
astronomers=20
do--pour a concrete slab block to attach their scope base, and surround =
the=20
block with several inches of sand---then there is no walkup tilt of the =
base,=20
and  any lateral mechanical vibrations at minimized as=20
well. 
   Well I =
have rambled=20
enough--good stability to all, and Season's Greetings!
 
          &nbs=
p;            =
;            =
            &=
nbsp;           &n=
bsp;         =20
Jim Lehman