In a message dated 12/06/01, bgoss@.................. writes:

>    I have read that a ball bearing was used in place of a knife edge in the 
> construction of the Lehman by some people.
>    I used the knife edge, but have problems with it when trying to adjust, 
> It seems to want to tilt then the boom does not have = swing in my case 
> east and west. In most of my data one side has just a bit more amplitude 
> than the other maybe 10% or less.

Dear Bryan Goss,

      There seem to be two satisfactory and practical solutions to the 
problem of providing a friction and offset free hinge on Lehman and other 
seismometers.  

      A ball bearing provides rolling contact rather than a knife edge. With 
a heavy weight it may be advisable to use a 1/4" dia or larger bearing. I 
used a special bit called a centre drill to drill into the end of the arm. 
The bit provides a small pilot hole and then a 90 deg cone, which is ideal 
for locating a ball bearing. You drill to just give the cone, degrease the 
hole and the bearing, apply epoxy to the cone and fill the centre hole, press 
the bearing into position and allow the epoxy to harden fully. However, a 
bearing is only as good as the surface it is sitting on. Using a polished 
bolt head has been recommended. However, you can provide an excellent surface 
by breaking a short length off a razor blade and sticking it to a bolt head 
with epoxy. I suggest that you machine / lap the bolt head flat, heat the 
bolt, apply the epoxy and clamp the joint firmly to give the minimum 
thickness glue line. The surface pressures under the ball are quite high.

      The other method is do away with the bearing surface entirely and use 
either a fine wire or a narrow strip of metal foil. The essential here is to 
firmly clamp the wire  where it leaves the joints. The wire must NOT enter a 
short Y section where it is not in contact with the clamp faces. The 8 thou 
dia. steel wire used on the header on many Lehmans is suitable. Of the two 
methods, this seems to require a bit more effort to fabricate. I understand 
that both methods will give stable pendulum periods in excess of 15 seconds, 
which is probably as much as the normal tilt sensitivity of the seismometer 
will allow. 

      Regards,

      Chris Chapman
In a message dated 12/06/01, bgoss@.................. writes:



  
; I have read that a ball bearing was used in place of a knife edge in the 

construction of the Lehman by some people.

   I used the knife edge, but have 
problems with it when trying to adjust, 

It seems to want to tilt then the boom does not have = swing in my case 

east and west. In most of my data one side has just a bit more amplitude 

than the other maybe 10% or less.




Dear Bryan Goss,



      There seem to be two satisfactory and practical solutions to the 

problem of providing a friction and offset free hinge on Lehman and other 

seismometers.  



      A ball bearing provides rolling contact rather than a knife edge. With 

a heavy weight it may be advisable to use a 1/4" dia or larger bearing. I 

used a special bit called a centre drill to drill into the end of the arm. 

The bit provides a small pilot hole and then a 90 deg cone, which is ideal 

for locating a ball bearing. You drill to just give the cone, degrease the 

hole and the bearing, apply epoxy to the cone and fill the centre hole, press 

the bearing into position and allow the epoxy to harden fully. However, a 

bearing is only as good as the surface it is sitting on. Using a polished 

bolt head has been recommended. However, you can provide an excellent surface 

by breaking a short length off a razor blade and sticking it to a bolt head 

with epoxy. I suggest that you machine / lap the bolt head flat, heat the 

bolt, apply the epoxy and clamp the joint firmly to give the minimum 

thickness glue line. The surface pressures under the ball are quite high.



      The other method is do away with the bearing surface entirely and use 

either a fine wire or a narrow strip of metal foil. The essential here is to 

firmly clamp the wire  where it leaves the joints. The wire must NOT enter a 

short Y section where it is not in contact with the clamp faces. The 8 thou 

dia. steel wire used on the header on many Lehmans is suitable. Of the two 

methods, this seems to require a bit more effort to fabricate. I understand 

that both methods will give stable pendulum periods in excess of 15 seconds, 

which is probably as much as the normal tilt sensitivity of the seismometer 

will allow. 



      Regards,



      Chris Chapman