thermal enhancements for separable thermal mechanical ... · thermal performance versus altitude....

Presented ByWilliam G. Anderson

Thermal Enhancements for Separable Thermal Mechanical Interfaces

James E. Schmidt, Matt Flannery, Jens Weyant Advanced Cooling Technologies, Inc.

Kevin Thorson, Lockheed Martin Advanced Technology Laboratories

Thermal & Fluids Analysis WorkshopTFAWS 2017August 21-25, 2017NASA Marshall Space Flight CenterHuntsville, AL

TFAWS Passive Thermal Paper Session

ISO9001 & AS 9100 CERTIFIED | ITAR REGISTERED

Agenda Background Information Improved Separable Thermal Mechanical Interfaces (STMIs) Thermal and Mechanical Testing Conclusion Future Work

2


Problem Statement

Many circuit card assemblies in current designs are temperature limited Heat is generated at the source (FPGA, processor, etc) and must be conducted to a chassis The card assembly is designed to be easily replaceable. As a result, a weak thermal link is presented by the mechanical clamp which allows the card to be swapped quickly Wedge lock clamps are preferred for this application but are poor thermal conductors

3

Standard Wedge Locks

Wedge Lock

Wedge Lock


Technical Approach COTS Wedge Lock Design Current commercially available card retainers only expand in one direction, limiting the heat paths from the card to the heat sink

Isothermal Card Edge (ICE‐Lok™) Design Reduces thermal resistance by increasing surface area for heat transfer and maximizing contact pressure. With two directions of expansion, the ICE‐Lok™ uses brackets to contact all four sides of the thermal interface. Integral Bracket reduces thermal resistance

4

Wedge Lock

ICE‐Lok with Integral Brackets


Thermal Enhancement: Spiral Lock and Ice‐Lok™

5

Current wedge lock designs Expands in one direction and has a high thermal resistance , which results

in two heat paths 80% of the power travels through the card‐chassis

interface, 20% of the power travels through the high thermal resistance wedgelock into the chassis (Q2)

Spiral Lock and ICE‐Lok™ Designs Expands in all directions Increases surface area contact between conduction card and chassis

thereby improving heat transfer with an overall lower thermal resistance

Current Wedgelocks

Spiral/ICE-Lok Concept

Chassis Wedge Lock

Conduction Card

Standard Wedge Lock Heat Transfer Path

80% 20%

Chassis Spiral/Ice‐Lok

Conduction Card

Heat Transfer Path


Requirements: VMEBus International Trade Association (VITA) VITA is a trade association with an associated ANSI recognized standards organization which has been developing practices and standards for VME and VPX system since 1983. It is comprised of leading industry experts and manufacturers.

Current card lock devices are made to integrate into both custom systems and systems made to ANSI/VITA specified chassis. The current chassis used by Lockheed Martin in tactical avionics programs meet these specifications.

ACT is using the VITA 48 standard for developing the geometry of the Spiral Lock to be used with 3U and 6U sized conduction cards.

6


Finalized Design Requirements Thermal Performance

Target specific thermal resistance of 0.10 deg C‐in/W Significantly reduce thermal resistance relative to COTS devices, (1.4 to 1.9 C‐in/W)

Integration Requirements Retain FOD if broken, one‐sided bolting, similar applied force/deflection to current devices.

Manufacturability Designed to VITA/ANSI specified chassis (VITA 48.0, 48.2, 48.5, 48.7) Target MRL 4 Select Type III, hard‐anodized, Class 2 (black) exterior surface finish for aircraft environment

Testing in Relevant Conditions Thermal Testing (Over Lifetime and at Altitude) Lifetime Testing (Mechanical Cycling of STMI) Retention Testing (New and at EOL) Shock and Vibration Vibration to MIL‐STD 883, conditions should encompass Atlas V, Delta IV, and Military Aircraft specs.

Shock to MIL‐STD 810, conditions should encompass Atlas V, Delta IV, and Military Aircraft specs.

7


Lockheed Martin Developmental Spiral Lock Bolting Mechanism: The Spiral Lock wedges were redesigned to allow a single actuating screw to pull the wedges.

8


Actuation/Chassis Deflection Requirements Actuation Mechanism: All current applications require the retainer to be actuated when only one side of the device is accessible. The original Spiral Lock design works best when actuated using two screws, with equal actuation from both sides.

Chassis Deflection: Current thermal chasses are not designed in anticipation of outwardly applied loading. Deflection of thin‐walled chassis, such as in liquid‐jacket cooled designs, could cause assembly issues or possible deformation.

9


Deflection Testing Chassis Deflection:

Current thermal chasses are not designed in anticipation of outwardly applied loading. Deflection of thin‐walled chassis, such as in liquid‐jacket cooled designs, could cause assembly issues or possible deformation.

Test Apparatus: A thin‐walled, easily deflected, chassis is utilized for mechanical characterization. Dial indicators and chassis are fixed to a stiff base plate. They can read outward movement of walls due to applied card retainer loads.

10

Deflection Testing Fixture Dial Indicator Positioning


Deflection Performance

Testing Parameters: Installation Torque: 10 in‐lbs (if possible)

Conclusions: The Spiral Lock will continuously deflect chassis walls until deformation occurs. Torque does not build due to elastic deformation of the chassis, so contact pressure cannot build. The COTS wedge lock works as expected, and produces nearly zero displacement.

11

‐0.01

0

0.01

0.02

0.03

0.04

0.05

Net Deflection [in

]

Spiral Lock (Actuation Redesign)

COTS Wedgelock (Baseline)


Friction Locking – Addressing Deflection “The Spiral Lock exerts a vertically applied, and a horizontally applied outward force upon its brackets; it is this latter component which causes deflection of the existing chassis side‐wall. Can change wedges to “Friction Lock” The vertically applied forces create friction between the bracket faces and the card‐guide/conduction card. If the friction forces developed by the wedges exceed the forces applied horizontally, by a critical ratio, then the card and the chassis cannot move relative to one another, locking the card to the chassis. By friction locking the card to the chassis, the card tab is used as a tensile member. This greatly stiffens the chassis.

12

Wedge Force Application

Bracket Friction Chassis Friction

Card Friction


ICE‐Lok™ (Integrated Card Edge) Friction Locking Wedges, Integrated Bracket Hi‐K Spiral Lock showed improved performance during Phase I program, identifying that more conductive brackets are of some value. Brackets integrated into wedges eliminate a contact resistance and essentially provide the same type of improvement offered by heat pipe integration.

13

ICE‐Lok


Qualification TestingNo industry standard qualification testing program currently exists for card lock devices. ACT and Lockheed Martin Performed Thermal, Lifetime, Retention, as well as Shock & Vibration Testing. Separate test chassis will be fabricated for the various testing tasks, including a dedicated test setup for retention and mechanical cycling. ACT’s ICE‐Lok™ designs will be tested in parallel with COTS Birtcherand Calmark card retainers. All devices installed side‐by‐side within the same test chasses. Throughout the life of each of the devices, thermal and mechanical testing data will be gathered for comparison.

14


Deflection and Retention Testing Deflection Prototypes were tested for deflection response of the thin‐walled chassis fixture (0.100” thickness) when tightened to 10 in‐lbs torque. The friction locking prototypes show similar deflection to COTS wedge locks, while the original Spiral Lock wedges continuously displace the chassis, never build 10 in‐lbs torque, and deform the chassis permanently if allowed to actuate.

Retention All devices were tested for card retention. All prototypes show no net deflection of a card after 300 lbs of force is applied.

15

‐0.01

0

0.01

0.02

0.03

0.04

0.05

Net Deflection [in

]

Net Chassis Deflection, 10 [in‐lbs]

ICE‐Lok Spiral Lock (Actuation Redesign)Wedgelock (35) (Coated) Wedge Lock (55) (Uncoated)COTS Wedgelock (Baseline) Wedge Lock (Birtcher)


Test Hardware

16

Linear Actuator

Mock Chassis

Card Displacement Sensor

Control Box and DAQ

Retention Testing Fixture

3‐Card Thermal Testing Fixture

Heater Block

Cold PlateConduction Card Card Guide


Thermal Qualification Testing A thermal test chassis for testing will be fabricated according the VITA/ANSI specifications

ICE‐Loks and COTS Wedge Locks will be installed in the test setup, and the test setup was installed in the thermal vacuum chamber.

The test setup operates operate over the considered altitude/air pressure range during each thermal test, providing a relationship of thermal performance versus altitude.

ACT will measure how the thermal test data changes as the STMIs are from new to their end‐of‐life service, 100 cycles.

Thermal resistance should not increase more than 0.7K/W

17

Thermal Testing Timeline

Thermal Test

Shock and Vibration

Thermal Test

20 Cycles

20 Cycles

20 Cycles

Thermal Test Chassis Fabrication Thermal

TestThermal Test

20Cycles

Thermal Test

20 Cycles

Thermal Test

Thermal Test

Thermal Test

Altitude [m]

Pressure [Torr]

Torque [in-lbf]

0 760 20

10000 196 20

20000 41 20


Thermal Resistance with respect to Bolt Torque

18

0.000

0.050

0.100

0.150

0.200

0.250

0.300

0.350

0 5 10 15 20 25

Thermal Resistance [°C/W]

Applied Torque [in‐lbf]

Spiral Lock (Original) ICE Lock COTS Wedge Lock (Baseline)

* COTS Wedge Lock is a Bircher 44‐5


Performance vs Elevation with Repeated Insertions

19

0.15

0.20

0.25

0.30

0.35

0.40

0 20 40 60 80 100 120

Thermal Resistance [C/W

]

Number of Cycles

ICE Lock (Group 1) (10000m) ICE Lock (Group 2) (10000m)

Calmark 280 (10000m) Birtcher 44‐5 (10000m)

0.15

0.20

0.25

0.30

0.35

0.40

0 20 40 60 80 100 120


]

Number of Cycles

ICE Lock (Group 1) (0m) ICE Lock (Group 2) (0m)Calmark 280 (0m) Birtcher 44‐5 (0m)

0.15

0.20

0.25

0.30

0.35

0.40

0 20 40 60 80 100 120


]

Number of Cycles

ICE Lock (Group 1) (20000m) ICE Lock (Group 2) (20000m)

Calmark 280 (20000m) Birtcher 44‐5 (20000m)

0.15

0.20

0.25

0.30

0.35

0.40

0 20 40 60 80 100 120


]

Number of Cycles

ICE Lock (Group 1) (Space) ICE Lock (Group 2) (Space)

Calmark 280 (Space) Birtcher 44‐5 (Space)

Resistance Increases With Elevation


Mechanical Qualification Testing Mechanical Testing Plan

Mechanical testing consisted of retention testing and shock and vibration testing. Conducted at Lockheed Martin Advanced Technology Laboratories Retention of the card under an applied for of 300lbf is necessary for qualification. ACT tested card retention on a dedicated chassis, then send the chassis for shock and vibration testing. Retention will be tested after shock and vibration. (MIL‐STD‐810G Shock, MIL‐STD‐883 Vibration)

Mechanical Testing Timeline The Spiral Locks are required to survive shock and vibration testing by retaining the card under load at beginning and EOL service.

Mechanical testing to verify retention with new STMIs installed into a VITA/ANSI compliant chassis. Retention tested before and after shock and vibration testing as per MIL‐STD‐810G and MIL‐STD‐883. Encompasses military aircraft, Delta IV, and Atlas V.

20

Mechanical Test Chassis

Fabrication

Retention Test

Shock and Vibration

Retention Test

Mechanical Testing Timeline


Shock and Vibration TestingNo problems observed with the ICE‐Lok™ The integrated bracket Spiral Lock showed consistent, adequate performance throughout the vibration tests, surviving through 100 mechanical cycles and eight hours of random vibration. Prototype after testing shown below

21


Conclusions and Recommendations The new ICE‐Lok™ card retainer is improved over other retainers currently on the market Redesigning the actuation mechanism and eliminating chassis deflection has succeeded in making ICE‐Lok a drop‐in, higher‐performance replacement for wedge lock type retainers. ICE‐Lok results in >10 C° lower component temperature at 100 W power input.

Testing of the ICE‐Lok™ over a 100 cycle lifetime, for both mechanical and thermal performance, in relevant environmental conditions has verified that the design of the retainer meets industry expectations of reliable performance in demanding environments Future Work Additional testing with ICE‐Lok™ and COTs STMIs Improve performance at high altitudes and in vacuum.

Need to look at coatings suitable for vacuum

22


Acknowledgements This Phase II program was sponsored by the Air Force Research Laboratory under Contract No. FA9453‐15‐C‐0418. Mr. Brenton Taft, Thermal Systems Lead, Air Force Research Laboratory/Space Vehicles was the Technical Monitor. We would like to thank him for his many helpful suggestions. The Laboratory Technicians at ACT were Philip Texter, Zachary Bitner, Phil Martin, and Larry Waltman. We would also like to acknowledge the help of Lockheed Martin engineers Greg Drexler and Nichole Murray who provided guidance and helped with the validation testing. The Lockheed Martin developmental Spiral Lock is covered by two patents (K. Thorson, R. Monson, et. al.): Cardlock clamp (8,559,178 dated 10/15/2013 and 8,743,544 dated 6/3/2014)

23

Thermal Enhancements for Separable Thermal Mechanical Interfaces

Jens Weyant, Lead [email protected]

Bill Anderson [email protected]

Thermal & Fluids Analysis WorkshopTFAWS 2017August 21-25, 2017NASA Marshall Space Flight CenterHuntsville, AL

TFAWS Passive Thermal Paper Session

thermal enhancements for separable thermal mechanical ... · thermal performance versus altitude....

Documents