In a message dated 25/01/2009 19:07:35 GMT Standard Time, nyponen@.........  
writes:

I'm a  newbie and have never built seismometer.  I'm wondering why
steel is  used so much in seismometer construction when other materials
have much  less thermal expansion.  Spruce wood, for example, has about
1/7 the  thermal expansion of steel.  Other materials expand even less,
quartz,  graphite, some glasses and ceramics, etc.  Some of the
materials with  low CTEs (Coefficient of Thermal Expansion) are
expensive (invar, tungsten,  polyimide, diamond), but others are
affordable.  Spruce is inexpensive  and widely available.  Here are a
couple of sources of graphite rods  and other  shapes:
http://www.waleapparatus.com/catalog.asp?prodid=548212&showprevnext=1
http://www.graphitestore.com/items_list.asp/action/prod/prd_id/25/cat_id/22



Hi Jim,
 
    Thermal expansion is not generally a serious  problem in seismometers, 
although you may need to allow for it. Wood is not used  for professional 
seismometers since it is sensitive to humidity and it tends to  warp. It also tends 
to take a 'set' under stress, particularly at elevated  temperatures. The 
thermal expansion is only small along the grain, otherwise it  may be quite large. 
It can be protected with adhesive + metal film,  but while helping, organic 
varnishes and paints do not completely exclude  moisture.
    Mild steel is readily available, has a high  rigidity, has the same 
properties in all directions and is not  expensive. But it does need to be 
protected from corrosion with one of the  'special' paints available. Do NOT use any 
ferromagnetic material for  the arm or the mass. It will pick up magnetic noise 
from the house wiring,  from the Earth and from passing trains, trucks and 
cars.
    Aluminum is more expensive and it is much less  rigid, but it is fairly 
resistant to corrosion. It can be varnished / painted  with advantage. Use only 
Stainless Steel, Al alloy or nylon bolts. Aluminum in  contact with mild 
steel corrodes rapidly. Brass is not much better. I have used  3" x 1" Al U 
channel sections for amateur type seismometers, but you can also  use Aluminum bar.  
Check _http://www.bgs.ac.uk/education/school_seismology/seismometer.html_ 
(http://www.bgs.ac.uk/education/school_seismology/seismometer.html) 
    Some of the materials you quoted are  GRAPHITE. Graphite is a LOW 
strength solid, NOT suitable for  constructing seismometers. You probably meant to 
refer to Carbon Fibre /  Monofilament. This is a very high strength material 
with a near zero expansion  coefficient along the filaments. The expansion 
perpendicular to the  filaments depends on the properties of the binding adhesive / 
resin and the  filling factor of the material. Both epoxy and vinyl adhesives 
may be used. The  properties of woven fabrics depend on the orientation of the 
filaments.
 
In general, the construction of seismometers needs  to be very rigid and 
moisture / corrosion resistant, whether this is achieved by  the material or by 
the sections used. I use brass for the mass. You can buy it  in a wide range of 
sections and sizes. I use a SS scalpel blade rolling on  a fixed 3/8" SS ball 
bearing for the bottom bearing on a Lehman, cost about  50c. You can also buy 
Tungsten Carbide needle rollers.

    Epoxy adhesives are very sensitive to surface  moisture and oil / grease. 
They don't stick soft metals like Al, Zn, Cu  etc. very well and most of the 
common resins cure brittle hard. The  flexible types may only be available in 
industrial packs and they may have  reduced strength. 
    Two component acrylic adhesives are tough, not  brittle and are much less 
effected by surface moisture or grease. I  use Holdtite ST3295. It is a thick 
liquid, but it mixes / spreads very  easily and sets quickly. It does not 
stick to polythene unless you oxidise the  surface. I prefer it to Devcon Plastic 
Weld or Bondloc Plastic Weld, which are  thick pastes. Loctite also make a 
wide range of these adhesives, as do Dow  Corning. 
 
    Regards,
 
    Chris Chapman






In a message dated 25/01/2009 19:07:35 GMT Standard Time, nyponen@gmail=
..com=20
writes:
<=
FONT=20
  style=3D"BACKGROUND-COLOR: transparent" face=3DArial color=3D#000000 size=
=3D2>I'm a=20
  newbie and have never built seismometer.  I'm wondering why
steel=20=
is=20
  used so much in seismometer construction when other materials
have much=
=20
  less thermal expansion.  Spruce wood, for example, has about
1/7 t=
he=20
  thermal expansion of steel.  Other materials expand even less,
qua=
rtz,=20
  graphite, some glasses and ceramics, etc.  Some of the
materials w=
ith=20
  low CTEs (Coefficient of Thermal Expansion) are
expensive (invar, tungs=
ten,=20
  polyimide, diamond), but others are
affordable.  Spruce is inexpen=
sive=20
  and widely available.  Here are a
couple of sources of graphite ro=
ds=20
  and other=20
  shapes:
http://www.waleapparatus.com/catalog.asp?prodid=3D548212&sh=
owprevnext=3D1
http://www.graphitestore.com/items_list.asp/action/prod/pr=
d_id/25/cat_id/22


Hi Jim,
 
    Thermal expansion is not generally a serious=20
problem in seismometers, although you may need to allow for it. Wood is not=20=
used=20
for professional seismometers since it is sensitive to humidity and it tends=
 to=20
warp. It also tends to take a 'set' under stress, particularly at eleva=
ted=20
temperatures. The thermal expansion is only small along the grain, otherwise=
 it=20
may be quite large. It can be protected with adhesive + metal film=
,=20
but while helping, organic varnishes and paints do not completely exclu=
de=20
moisture.
    Mild steel is readily available, has a high=20
rigidity, has the same properties in all directions and is not=20
expensive. But it does need to be protected from corrosion with one of the=20
'special' paints available. Do NOT use any ferromagnetic material for=20
the arm or the mass. It will pick up magnetic noise from the house wiri=
ng,=20
from the Earth and from passing trains, trucks and cars.
    Aluminum is more expensive and it is much=20=
less=20
rigid, but it is fairly resistant to corrosion. It can be varnished / painte=
d=20
with advantage. Use only Stainless Steel, Al alloy or nylon bolts. Aluminum=20=
in=20
contact with mild steel corrodes rapidly. Brass is not much better. I have u=
sed=20
3" x 1" Al U channel sections for amateur type seismometers, but you can als=
o=20
use Aluminum bar.  Check h=
ttp://www.bgs.ac.uk/education/school_seismology/seismometer.html
    Some of the materials you quoted are=20
GRAPHITE. Graphite is a LOW strength solid, NOT suitable for=20
constructing seismometers. You probably meant to refer to Carbon Fibre=20=
/=20
Monofilament. This is a very high strength material with a near zero expansi=
on=20
coefficient along the filaments. The expansion perpendicular to the=20
filaments depends on the properties of the binding adhesive / resin and the=20
filling factor of the material. Both epoxy and vinyl adhesives may be used.=20=
The=20
properties of woven fabrics depend on the orientation of the filaments.

    In general, the construction of seismometers ne=
eds=20
to be very rigid and moisture / corrosion resistant, whether this is achieve=
d by=20
the material or by the sections used. I use brass for the mass. You can buy=20=
it=20
in a wide range of sections and sizes. I use a SS scalpel blade rolling =
;on=20
a fixed 3/8" SS ball bearing for the bottom bearing on a Lehman, cost a=
bout=20
50c. You can also buy Tungsten Carbide needle rollers.
    Epoxy adhesives are very sensitive to surface=20
moisture and oil / grease. They don't stick soft metals like Al, Zn, Cu=
=20
etc. very well and most of the common resins cure brittle hard. Th=
e=20
flexible types may only be available in industrial packs and they may have=20
reduced strength. 
    Two component acrylic adhesives are tough, not=20
brittle and are much less effected by surface moisture or grease. =
I=20
use Holdtite ST3295. It is a thick liquid, but it mixes / spreads very=20
easily and sets quickly. It does not stick to polythene unless you oxidise t=
he=20
surface. I prefer it to Devcon Plastic Weld or Bondloc Plastic Weld, which a=
re=20
thick pastes. Loctite also make a wide range of these adhesives, as do Dow=20
Corning. 
 
    Regards,
 
    Chris Chapman