Engineering Department ENEN Screwed solutions for the new tertiary collimator jaw assembly Federico Carra (CERN EN/MME) Geneva, 4 th September 2014 4th.

Engi

neer

ing

Depa

rtm

entEN

F. Carra– CERN / EN-MME 1

Screwed solutions for the new tertiary collimator jaw assembly

Federico Carra (CERN EN/MME)

Geneva, 4th September 2014

4th September 2014

Engi

neer

ing

Depa

rtm

entEN

F. Carra– CERN / EN-MME 2

Engineering in the MME design office

Example of screwed assembly design: LHC tertiary collimator

Outlook

4th September 2014

Engi

neer

ing

Depa

rtm

entEN

F. Carra– CERN / EN-MME 3

10 Engineers in MME design office (mechanical, aeronautical, materials, …)

Project coordination

Calculations in pre-design and design phases

Analyses on components already installed in the accelerators

Diagnostics on device malfunctioning

R&D: materials, test organization and design, etc.

Engineering specifications, safety files, safety procedures

Engineering at MME design office

4th September 2014

Engi

neer

ing

Depa

rtm

entEN

F. Carra– CERN / EN-MME 4

Engineering in the MME design office

Example of screwed assembly design: LHC tertiary collimator

Outlook

4th September 2014

Engi

neer

ing

Depa

rtm

entEN

F. Carra– CERN / EN-MME 5

Beam-induced accidents represent one of the most dangerous and though less explored events for Accelerators.

Beam Intercepting Devices (BID) are inherently exposed to such events

LHC beam energy is 2 orders of magnitude above previous machines. Stored energy density is 3 orders of magnitude higher.

Novel, yet-to-characterize, composite materials are under development to meet these challenges.

New sophisticated and powerful Numerical Tools (Hydrocodes) are used to simulate accidental events.

4th September 2014

LHC Collimators: Context

Engi

neer

ing

Depa

rtm

entEN

64th September 2014 F. Carra– CERN / EN-MME

Vacuum tank

Jaws

Actuation system

Jaws 40µm surface flatness on 1m 10µm positioning accuracy Heat load: up to 30 Kw

Particle Beam

Engi

neer

ing

Depa

rtm

entEN

F. Carra– CERN / EN-MME 7

Tungsten(Inermet® IT180)

TCTP collimator: brazed vs. screwed

4th September 2014

50 M4 stainless steel screws (A4-100 silver-coated) are used to fix tungsten blocks to the housing

2 options for the assembly housing/cooling circuit/stiffener:

Brazing with a silver alloy

Screwing with 54 M4 screws (A4-100 silver coated)

All the calculations of tightening torque and preload are done according the German Standard VDI2230

Brazing

Engi

neer

ing

Depa

rtm

entEN

F. Carra– CERN / EN-MME 8

TCTP collimator: brazed vs. screwed

4th September 2014

BRAZED Good thermal contact between the

components

No risk of relaxation at the interfaces in operation

Expensive and complicated procedure

SCREWED Simple and quick assembling

Cheap

Thermal contact dependent on the screw preload

Risk of stress relaxations at the interfaces, especially during bake-out cycles at 250 ˚C

Engi

neer

ing

Depa

rtm

entEN

F. Carra– CERN / EN-MME 9

Thermal contact at the interfaces

4th September 2014

In a screwed contact, the thermal conductance is a function of the pressure at the interface:

The higher the conductance, the lower the jaw maximum temperature in operation and the thermally-induced deformations!

With the pressure given by the M4 screws, the calculated conductance is 7000 W/m2K

A thermo-mechanical finite element analysis was necessary to understand if the conductance granted by the screws was high enough to guarantee temperatures and deformations within the component specification.

Thermal conductanceContact pressure

Engi

neer

ing

Depa

rtm

entEN

Temperatures

Deformations

• Conductance 7000 W/m2K

• Initial temperature: 27 ºC

• Flow rate inside cooling pipes: 5 l/min

• Bulk temperature of cooling water: 27 ºC

• Thermally induced deflection ~ 50 m;

• Deflection provoked by gravity ~ 25 m;

• Total deflection ~ 50 + 25 (worst case) = 75 m

• Specification: 100 mm OK!

Thermo-mechanical behaviour

The conductance theoretically achievable with the screws

(7000 W/m2 K) would be enough to guarantee temperatures

and deformations during operation within the specification

Engi

neer

ing

Depa

rtm

entEN

Experimental conductance measurement

The conductance calculation was validated with experimental measurements performed by I. Leitao with an ad-hoc setup

The prototype produced contained 1 screwed jaw and 1 brazed

The screwed jaw has an acceptable conductance in every measured section

Surprisingly, the brazed jaw has a conductance not so much higher than the screwed one!

Engi

neer

ing

Depa

rtm

entEN

F. Carra– CERN / EN-MME 12

Could creep-induced deformations occur during bake-out (~ 250 ºC)?

Eq. stress on CuNi < 60 MPa

Eq. stress on Glidcop Al-15 < 100 MPa

No significant creep deformations expected on TCTP jaw components

Copper OFE

Copper OFS

CuNi alloys

Glidcop

Creep Resistance

Low

High

Zoom on Glidcop stiffener: stress under the head of a screwOrange area: equivalent stress comprised between 60 and 100 MPa

Stress relaxation: FEM analysis

4th September 2014

Engi

neer

ing

Depa

rtm

entEN

F. Carra– CERN / EN-MME 13

In case creep and stress relaxation occur, a decreasing in conductance is expected

To rule out creep, we performed complete 3 bake-out cycles (~6days at 250 ˚C), measuring the thermal conductance between pipes and jaw before and after the treatment

Stress relaxation: experimental validation

4th September 2014

No negative effect has been observed!

Engi

neer

ing

Depa

rtm

entEN

F. Carra– CERN / EN-MME 144th September 2014

Conclusions

Two different jaw designs were proposed for the new tertiary collimators: brazed

and screwed

Numerical and analytical calculations, followed by experimental measurements,

ruled out problems of unsufficient thermal conductance and creep resistance

of the screwed assembly

The prototype produced features one brazed jaw and one screwed, and could

potentially be installed in the LHC to prove the validity of the two solutions

Engi

neer

ing

Depa

rtm

entEN

F. Carra– CERN / EN-MME 15

Depa

rtm

ent o

f Mec

hani

cal a

nd

Aero

spac

e En

gine

erin

g (D

IMEA

S)

Thank you for your attention!

4th September 2014

Engi

neer

ing

Depa

rtm

entEN

F. Carra– CERN / EN-MME 164th September 2014

Conclusions

MME design office engineers are involved in several phases of a component life (pre-design, design, installation, operation).

Analytical and/or empirical techniques are often adopted when studying problems that can be simplified to a certain extent.

For complex structures, implicit FEM codes like ANSYS are usually adopted to perform structural analyses (static, transient, modal, harmonic, …) as well as multiphysics calculations (e.g. thermo-structural, electro-thermal, etc.).

In some exceptional cases, involving material explosion, spallation, change of phase, high non-linearities, the study is performed with Hydrocodes like Autodyn or LSDyna.

An intense material R&D program is also pursued, in collaboration with other MME sections (MME/MM, mechanical laboratory, mechanical workshop), different CERN groups and departments, as well as external European partners, mainly in the frame of the Eucard collaboration (PoliTo, Kurchatov Institute, GSI, RHP, BrevettiBizz, EPFL, etc.).

Engi

neer

ing

Depa

rtm

entEN

F. Carra– CERN / EN-MME 17

Prim

ary

Colli

mat

or p

e

p

Core

Unavoidable losses

Shower

Beam propagation

Impact parameter

≤ 1 mm

Primary halo (p)

p

C-C Collimators are affected by intrinsic limitations which may ultimately limit LHC performances:• Poor electrical conductivity (High RF impedance)• Limited Radiation Hardness (Reduced Lifetime)• Low-Z material (Limited Cleaning Efficiency)

Abs

orbe

r

C-C C-C W W

Super-conducting

magnets

SC magnets and particle physics

exp.Courtesy: R. Assmann

Novel Materials for Collimators

Abs

orbe

r

Phas

e II

Co

llim

ator

Seco

ndar

y Co

llim

ator

The collimation system must satisfy 2 main functions:

• Multi-stage Beam Cleaning, i.e. removing stray particles which would induce quenches in SC magnets.

• Machine Protection, i.e. shielding the other machine components from the catastrophic consequences of beam orbit errors.

4th September 2014