Temperature sensors (as draft for scope review) A. Cervera M.A. - - PowerPoint PPT Presentation

Temperature sensors

(as “draft” for scope review)

• A. Cervera

M.A. Garcia Peris (IFIC-Valencia)

CISC, DUNE CM FNAL, 06/06/2019

Anselmo Cervera Villanueva, IFIC-Valencia

Table of contents

2

# slides accum. # slides Content 1

1 Motivation of standard temperature measurements

5

4 Motivation of precision temperature measurements, CFD simulations and overall strategy

9

4 Short description of all systems in ProtoDUNE-SP

12

3 Short description of laboratory calibration and results

14

2 Strategy for extrapolation from ProtoDUNE to DUNE

18

4 Short description of all systems in DUNE-SP with lessons learned from ProtoDUNE-SP . Baseline number and location of devices

23

5 Some details of static precision thermometers (profilers and individual) to demonstrate feasibility and cost effectiveness: mechanics, cable routing, etc.

43

20 Charge questions and answers: Here is where most of the plots will go

Anselmo Cervera Villanueva, IFIC-Valencia

Draft charge

1.How did the three types of temperature sensors employed in the ProtoDUNE detectors (Static T-gradient thermometer, Dynamic T-gradient thermometer, and individual temperature sensors) perform relative to one another? 2.Were the employed temperature sensors able to provide the data necessary to validate theoretical calculations of fluid flow inside the ProtoDUNE cryostat? 3.Is it important from a physics standpoint to validate theoretical calculations of fluid flow inside the DUNE far detector modules (no longer a design issue) and if so what level of instrumentation will be required to accomplish this? 4.Are all three different types of temperature monitors employed in the ProtoDUNE-SP cryostat necessary for the DUNE far detector cryostats in that they have unique features necessary for meeting monitoring requirements? 5.Are the temperature monitor strings currently being implemented in the ProtoDUNE-DP cryostat (and hence not yet tested) potentially a simpler and less expensive option to the three types of devices implemented in the ProtoDUNE-SP cryostat? 6.Are the proposed mechanisms for supporting the monitors within the cryostat and connecting them to the outside of cryostat mechanically sound and cost effective?

3

Anselmo Cervera Villanueva, IFIC-Valencia

Question 1

• How did the three types of temperature sensors employed in the

ProtoDUNE detectors (Static T-gradient thermometer, Dynamic T- gradient thermometer, and individual temperature sensors) perform relative to one another?

• The dynamic T-Gradient monitor has been used to validate the

hypothesis of uniform temperature when pumps are off, used by all

• ther sensors for the so called pumps-off calibration
• The dynamic and static profiles can be compared at any time to

study the differences between the two locations

• The 2D sensor array at bottom at top can be used to connect the

measurements between the two profilers and also to understand horizontal effects due to the location of LAr inlets and LAr pump

4

Anselmo Cervera Villanueva, IFIC-Valencia

5 dynamic static dynamic static

Anselmo Cervera Villanueva, IFIC-Valencia

5 dynamic static dynamic static

Anselmo Cervera Villanueva, IFIC-Valencia

5 dynamic static dynamic static

Anselmo Cervera Villanueva, IFIC-Valencia

Question 2

• Were the employed temperature sensors able to provide the data

necessary to validate theoretical calculations of fluid flow inside the ProtoDUNE cryostat?

• We are still in the process of tuning CFD simulations using data from

ProtoDUNE-SP data

• We expect to have some systematic studies before the review
• Current simulations predict 5-10 mK gradients, which agree with

data to first order, but we need to go into the details

6

Anselmo Cervera Villanueva, IFIC-Valencia

Current simulations

7

PUMPS ON PUMPS OFF

https://indico.fnal.gov/event/20887/contribution/0/material/slides/0.pdf

Anselmo Cervera Villanueva, IFIC-Valencia

Next steps

• The plan is to identify the key parameters and make several

simulations varying one parameter at the time:

• LAr surface temperature
• LAr inlet temperature

8

Parameter Input or Output of CFD Version 1 Comments on Version 1 Cryostat height Input 7.878 m Measured with laser (1 cm error approx.) LAr surface height Input 7.406 m From capacitive level meter: https://np04-slow-control.web.cern.ch/ np04-slow-control/app/#!/histogram/NP04_DCS_01_NP04_LT0100 ullage pressure Input 1.045 bar https://np04-slow-control.web.cern.ch/np04-slow-control/app/#!/ histogram/NP04_DCS_01_NP04_4PT4920 LAr surface temperature Input 87.596 K Computed using https://lar.bnl.gov/properties/basic.html#phase and the ullage pressure. Compatible with temperature measured by sensor Heat flux Output 5.76 W/m^2 LAr inlet temperature Input

• utlet temp + 0.2 K Estimated by M.Chalifour

LAr flow rate per pipe Input 0.4170025 kg/s vapor being drawn from the chimneys Output 5-7 gr/sec Estimated by M.Chalifour.

table not up to date current inputs

Anselmo Cervera Villanueva, IFIC-Valencia

Question 3

• Is it important from a physics standpoint to validate theoretical

calculations of fluid flow inside the DUNE far detector modules (no longer a design issue) and if so what level of instrumentation will be required to accomplish this?

• Energy calibration depends very much on the local electron lifetime
• Give some numbers
• Electron lifetime can only be measured in few locations near the

cryostat corners by Purity Monitors. This is an instantaneous measurement

• Electron lifetime can be computed using cosmic rays, but those are

rare 1500 m underground:

• Need here some numbers: cosmic muon flux, statistic needed

for reasonable understanding and correction, etc

9

Anselmo Cervera Villanueva, IFIC-Valencia

Question 3

• CFD simulations, tuned with temperature data will be used to

predict the electron lifetime everywhere in the TPC

• In principle the more sensors the better constraint, which

implies a lower systematic error in the energy correction

• There is obviously saturation here. We are still far from

understanding how many sensors we need

• To understand the optimal number of sensors and their locations we

need to implement the full correction chain:

• 1. several CFD simulations varying relevant parameters
• 2. mock temperature data with varying number of sensors to constraint the CFD

simulations

• 3. Estimate the error on the predicted electron lifetime, which will depend on the

number of sensors and their location

10

Anselmo Cervera Villanueva, IFIC-Valencia

Question 4

• Are all three different types of temperature monitors employed in the

ProtoDUNE-SP cryostat necessary for the DUNE far detector cryostats in that they have unique features necessary for meeting monitoring requirements?

• Dynamic T-gradient monitor: can be calibrated in-situ at any time,

ensuring that we have a reliable understanding of the gradient regardless of ageing effects, calibration or electronics problems

• Static T-gradient monitor: It has much simpler mechanics. It can

be installed in places were space is restricted. In DUNE-SP , where there is no space behind APAs, this is the only viable solution for the vertical gradient measurement. Lab calibration works since T0

• Individual sensors: They complement the vertical measurement

with a 2D horizontal map. In ProtoDUNE-SP it has been observed that the horizontal behaviour is far from trivial

11

Anselmo Cervera Villanueva, IFIC-Valencia

Pipes sensors

• Plots in the next slides show the difference in temperature between

the two pipe edge sensors

12 dynamic static

Anselmo Cervera Villanueva, IFIC-Valencia

Pipe 1 (from TCO)

13

Anselmo Cervera Villanueva, IFIC-Valencia

Pipe 2

14

Anselmo Cervera Villanueva, IFIC-Valencia

Pipe 3

15

Anselmo Cervera Villanueva, IFIC-Valencia

Pipe 4

16

Anselmo Cervera Villanueva, IFIC-Valencia

Far from inlets

17

Anselmo Cervera Villanueva, IFIC-Valencia

Pumps-off for all sensors

• Second set of 6 sensors (one SUB-D connector) shows larger RMS
• First set within 5 mk

18

3 mK 2 mK 5 mK 5 mK

Anselmo Cervera Villanueva, IFIC-Valencia

Pumps-on

• Pipes sensors in October

19

• Pipes sensors in December

Anselmo Cervera Villanueva, IFIC-Valencia

Pumps-on

20

• Pipes sensors in Febrero

Anselmo Cervera Villanueva, IFIC-Valencia

Pumps-on

21

• Pipes sensors in Abril

Anselmo Cervera Villanueva, IFIC-Valencia

Pumps-on

22

Anselmo Cervera Villanueva, IFIC-Valencia

Question 5

• Are the temperature monitor strings currently being implemented in

the ProtoDUNE-DP cryostat (and hence not yet tested) potentially a simpler and less expensive option to the three types of devices implemented in the ProtoDUNE-SP cryostat?

• NO, ProtoDUNE-DP uses one string in one of the corners with

standard uncalibrated sensors. In principle this can only be used for level measurements and during cooldown and filling. Those sensors don’t have sufficient precision for:

• Detecting cryogenic system problems: < 20 mK (in the 35t

stratification was observed with 20 mK gradient)

• Check the temperature uniformity: in ProtoDUNE-SP <10 mK
• Predict the electron lifetime based on CFD simulations
• This system can only be installed in the corners

23

Anselmo Cervera Villanueva, IFIC-Valencia

Question 6

• Are the proposed mechanisms for supporting the monitors within

the cryostat and connecting them to the outside of cryostat mechanically sound and cost effective?

• Static T-Gradient monitors:
• The proposed mechanics is much cheaper that it was in

ProtoDUNE-SP since no Faraday Cage is needed

• The installation is also simpler since no rigid elements are used

and can be transported in a roll to the cryostat

• Individual sensors:
• Mechanics will be simpler than in ProtoDUNE-SP

. Cables will be routed in bundles

• In both cases cost is driven by sensors & cables, not by the

supporting system

24