Infrared Detection of Defects in Green-State and Sintered PM - - PowerPoint PPT Presentation

1

Infrared Detection of Defects in Green-State and Sintered PM Compacts

Souheil Benzerrouk

Worcester Polytechnic Institute October 27-28, 2004

Morris Boorky Powder Metallurgy Research Center

Reinhold Ludwig

2

A warm Welcome from Prof. Ludwig

3

Research Objectives

 Evaluate the feasibility of IR imaging for the

detection of surface and subsurface defects in P/M parts

 Establish a full dynamic thermo-electric IR

solution that allows the testing of both green state and sintered P/M parts

 Estimate experimentally effects from

equipment radiation in a manufacturing environment

4

Research Approach

 Establish the necessary background in the fields of

radiation, imaging and detection

 Construct a dynamic test bed to test for subsurface

defects

 Test Controlled samples with subsurface defects  Process evaluation through on-line testing of P/M

parts

5

Outline

 IR imaging: sources of radiation  Test Arrangement  Experimental study: subsurface defects imaging and

data processing

 Experimental Study: on-line testing of green state

parts

 Accomplishments  Future work

6

IR imaging

 Sources of radiation



Incident from the sample



Reflected



Atmospheric

at r s

S S S S + + =

Contacts Sample under test IR camera Signal processing computer DC power supply

Incident radiation from the surroundings Emitted radiation Reflected radiation Radiation from the surroundings

Generic Formulation

7

Test Arrangement

 DC Power supply  Motorized press

system

 An IR camera at

0.3m away from the sample under test

 A computer for

processing and camera controls

Test Arrangement

Contacts Sample under test IR camera

Firewire

Signal processing computer DC power supply

GPIB

Switch Function generator

8

Test Arrangement

 A switching circuit,

for pulse shaping and synchronous

• peration

Test Arrangement

Press system Painted P/M part IR camera Switching circuit Control computer DC Power supply

9

Subsurface Defects

 Part parameters:

 Green state part with

pure iron: 1000B

 No lubrication

 Defect:

 Location:2 mm from

the surface, 2.5 cm from the top

 1 mm hole

Experimental Results

Current step

10

Subsurface Defects: Processing

Experimental Results

11

Experimental Results

Subsurface Defects: Processing

Spot Temperature Over Time

302 302.2 302.4 302.6 302.8 303 303.2 303.4 5 10 15 20 Temperature (Kelvin)

Signature from the subsurface defect

12

Transient Model: Reminder

Current step

t =0.2 sec t =1 sec

2D Study and Sensitivity Estimation

13

Transient Model: Reminder

 Surface and subsurface

defects are easily detected

 Fast response, very

efficient, suitable for a go/no go test

 Highly sensitive, reduced

post-processing complexity

 Camera requirements

include:

 Dynamic range: 2 Hz  Thermal sensitivity: 0.2 0C

2D Study and Sensitivity Estimation

14



Part Parameters:



Density: 7.2 g/cc



Material: FC0205



Lubricant: 0.55% EBS



Embedded Defects:



Material: Glass bead



Size: 0.8mm

Experimental Results

Subsurface Defects: New Samples

Parts courtesy of Nichols Portland

15

Experimental Results

Subsurface Defects: New Samples

Parts courtesy of Nichols Portland

16

Experimental Results

Subsurface Defects: Discussion

 Method successful in the detection of subsurface

defects

 Thermal signature is dependent on defect size,

shape, orientation and distance from the surface: Diffusion

 Smaller defects or deeply imbedded defects

(distance from the surface is greater than the size of the defect) require a stable background and a sensitive detector

17

Experimental Results

On-Line Testing

 Requirements:

 Fast response  High spatial resolution  High temperature range  High image recording rate

 Benefits:

 Allows 100% testing  Provides real time feedback on part quality and

process repeatability

18

On-Line Testing

Parts courtesy of GKN Worcester

Experimental Results

 Part Constituents:

 FLC-4608  0.9% graphite content  0.75% KENOLUBE P-11

Lubricant

 Manufacturing rate:

500parts/hour

19

On-Line Testing

Setup courtesy of GKN Worcester

Experimental Results

20

On-Line Testing: Processing

 Processing of individual

parts

 Profiles have very similar

shapes

 Some difference due to

Part1 having a different angle when manufactured

Experimental Results

21

On-Line Testing: Processing

Experimental Results

300 305 310 315 320 325 330 335 340 345 5 10 15 20 25 30 35

Time (sec) Temperature (Degree K)

Monitoring the temperature of a spot in the production line

22

On-Line Testing: Processing

Experimental Results

Time (sec) Temperature (Degree K)

Monitoring the temperature of a spot in the production line and zooming in one part

23

On-Line Testing: Discussion

 After processing valuable information about the part

integrity and quality is available in real time

 Process variations causing density gradients can be

flagged

 Part orientation in the line can be detected  Easy to implement with low cost and space overhead

Experimental Results

24

Accomplishments

 Developed an analytical foundation of heating with

direct current (electrostatics and heat transfer) and IR detection

 Built a suitable model for predicting the thermal profile

• n the surface of a part ( close to what the IR camera

will capture)

 Conducted simple dynamic testing of controlled

samples with subsurface defects

 Conducted on-line testing of simple parts

Accomplishments

25

Future Work

 Experimental measurements (Controlled samples)

 Different powder mixture  Different compaction densities  Different lubricants and concentrations  Glass and plastic inserts to simulate subsurface defects

 Instrumentation efforts

 Current strength and pulse shape  Injection methods  Post processing options

 Include in the numerical model

 Radiation computations  Density variation and non-uniformity  Contact resistance (material parameters)

Future Work

26

Questions

27

Questions