Post on 11-May-2015
description
A measurement of the Planck constant using pressure metrology
C M SuttonMeasurement Standards Laboratory of New Zealand
Industrial Research Limited
v1014f
Acknowledgement: Many people - MSL, IRL, overseas co llaborators …
NZIP Conference, Wellington, 17 - 19 October 2011
2
Contents
� Introduction� The Planck constant & how it relates to the kilogra m
�Watt balance� How it works
�A pressure balance watt balance� Concept� How it compares with other watt balances� Research activities
�Summary
3
Introduction
�Why measure the Planck constant?
� Because the present artefact kilogram is limiting development of the SI� The International System of Units
� A new definition of the kilogram in terms of the Planck constant is a way forward
and
International prototype kilogram( IPK)
4
What is the problem with the kilogram?
�The present kilogram is the mass of the IPK- International Prototype kilogram� Only reliable to ~ 50 µg or 5 parts in 10 8 - or worse?� Limiting other units� ampere� mole� candela
Apparent variations in masswith time for copies of the IPK
In particular: The ampere could be defined in terms of quantum phenomena� Josephson volt� Quantum Hall effectand realised with a relativestandard uncertainty uR ~ 10-9
5
� Link kg to NA� Counting atoms in sphere
of single crystal 28Si� International collaboration
- Australia, France, Germany,Italy,Japan, USA, …
� Budget several M€!
Solution - re-define the kilogram
� .. In terms of a fundamental constant� Planck constant h or Avogadro constant NA
� Only current options likely to achieve accuracy uR < 5 ×××× 10-8
� Link kg to h� Watt balance - Kibble 1975
� USA, Canada� France (2),
Switzerland,China,New Zealand …
6
Current results for h & NA
� Lowest uR = 3.0 ×××× 10-8
� BUT span of results is 31 ×××× 10-8
� More measurements needed !
6.022138
6.022139
6.022140
6.022141
6.022142
6.022143
CODATA 2006 NIST wb 2007 NPL wb 2007 IAC 2010
Avo
gadr
o co
nsta
nt /(
1023
mol
-1) Target accuracy:
uR ≈≈≈≈ 2 ×××× 10-8
( ) 20
A2rcA e M
hR N
α
∞
=
Challenging:~2 s in 3 years or ~2 mm in 100 km!
USAwatt balance
UKwatt balance
Silicon sphere
7
A watt balance - how it works
� Mechanical versus electrical power - two modes
Dynamic mode
Weighing mode
γ= × = ⋅ =∫uurr r r r
F I dl B I m g
~ coil length dl
( ) ( )U B dl dl B v vν γ= × ⋅ = − × ⋅ = − ⋅∫ ∫uur uur rr rr r
Factor γγγγ = U/v from dynamic mode
� When I & U are measured– with quantum electrical standards
» Josephson volt, quantum Hall resistance
m g U
I vγ = = and
2JfU I
m C hg v gv
= =Hence
8
Gas pressure balance:• No piston-cylinder contact
- aerodynamic bearing- strong piston centring forces
• Small piston-cylinder gap- < 1 µµµµm
Concept - pressure balance watt balance
�Weighing mode� Two pressure balances
� As force comparator� Coil fixed on piston
�Dynamic mode� Oscillatory coil motion
� ~ 1 Hz, noise rejection
Aim: Table-top size watt balance
9
What do other watt balances look like?
�Quite different!�Traditional mass/force balance
� Coil hanging from gimbals
�Various means to move coil� Beam, wheel, flexures …
� Un-wanted forces, motions� Need to control coil position
– all six degrees of freedom
�Constant coil velocity� In dynamic mode
� dc induced voltagemeasurement
� Susceptible to noise
USA watt balance
10
Research - Pressure balances 1
�Weighing performance� MSL twin pressure balance 2
� Repeatability vs load, (AB)4 or (AB)5 loading sequence– A ~ unloaded, B ~ loaded, Calculate dp and u(dp)
» DHI pressure balances
� Near zero load u(dp) ~ 0.25 mPa or 2.5 ×××× 10-9 of line pressure
0.0
0.2
0.4
0.6
0.8
0 50 100 150 200 250
Mass /g
u(d
p) /
mP
a Nearly good enough
11
Research - Pressure balances 2
� Investigating� Damped resonant behaviour
� To improve short-term repeatability
� Damping depends on� pressure, gas, geometry� Due to non-adiabatic behaviour of gas
– With NIST Gaithersburg& DH Instruments, USA
-0.75
-0.5
-0.25
0
0.25
0.5
0.75
0 20 40 60
Time /s
Am
plitu
de /m
m0
5
10
15
20
0 1 2 3
Period t 0 /s
Q =
ππ ππ/ λ
λ λ λ
CEC
DHI
Aim toreduce damping
& increase Q
12
Research - Pressure balances 3
�New pressure balance design� With IRL Mechanical workshop
� Stationary piston– Allows wires to coil on piston
� Rotating cylinder� Trial design for cylinder rotation
– Axis defined by matched pair ofangular contact ball bearings
� Prototype made– Currently being tested
13
Research - Ground vibration at IRL
�Noise rejection� Choose oscillation frequency - for dynamic mode
�Where background noise is low
� Fourier analysis of d & U - to give γγγγ�Rejects noise at other frequencies
� Ground vibration�Low from 1 Hz to 3 Hz
� Matches preferredfrequency range�0.1 Hz to 5 Hz
–To avoidmechanicalresonances
1.E-10
1.E-09
1.E-08
1.E-07
1.E-06
1.E-05
1.E-04
0.1 1 10 100Frequency /Hz
Leve
l of v
ibra
tion
Amplitude /m
Velocity /(m/s)
Acceleration /(m/s 2)
14
Research - Dynamic coil position measurement
�High-speed heterodyne laser interferometry� For measuring oscillatory coil motion� ~ 1 MHz sample rate
� Developing processing electronics� With Intelligent Machines & Devices Group
-2.5
-1.5
-0.5
0.5
1.5
2.5
550 650 750 850 950
Time /s
App
aren
t dis
plac
emen
t /nm
15
New cryo-cooled cryostat
Research - Induced voltage measurement
�Via ac Josephson voltage standard� Have purchased cryo-cooler� Plan to purchase PJVS
� Programmable Josephsonvoltage standard
� With NIST Boulder, USA
� Investigating� Differential sampling voltmeter
� To measure the difference betweenthe induced coil voltage and theac Josephson voltage
� With NIST Gaithersburg, USA
16
0.9520
0.9522
0.9524
0.9526
0.9528
0.9530
0.9532
0 10 20 30 40 50 60 70 80 90
Distance z in air gap /mm
Rad
ial i
nduc
tion
B (
LNE
) /T
esla
0.9299
0.9300
0.9301
0.9302
0.9303
0.9304
0.9305
0.9306
Rad
ial i
nduc
tion
B (
MS
L) /T
esla
LNE calculationMSL calculation
Research - Magnetic field
�Trial calculations on magnet design� With HTS110� Obtained similar variations
� Of B with z to lab in France
�Rig to measure RTC� Reversible temperature
coefficient� Of permanent magnets� Want ~zero RTC� With Electron Energy, USA
17
Summary
� Pressure balance watt balance� Concept established
� Significantly different from existing watt balances
� Research initiated on key factors� Influencing feasibility and performance
� Aiming for operational watt balance� In some form, mid-2013� Results in advance of Dec 2014 - CODATA
Any questions?