Computational Estimation of Heat Transfer Curves for Microstructure Prediction and Decision Support

1
Computational Estimation of Heat Transfer Curves
for Microstructure Prediction and Decision Support

• Aparna S. Varde, Mohammed Maniruzzaman, Elke A.
  Rundensteiner and Richard D. Sisson Jr.
• Worcester Polytechnic Institute
• Worcester, MA, USA.
• Presented at TMS-04, New Orleans, Louisiana, USA.

2
Introduction

• The result of a Heat Treating experiment is often
  represented as a graph called a Heat Transfer
  Curve, i.e.,
• Heat Transfer Coefficient as a function of
  temperature.
• Performing lab experiment consumes time and
  resources.
• Desirable to computationally estimate the Heat
  Transfer Curve, given experimental input
  conditions.
• Estimation can be used for various applications,
  e.g.,
• Microstructure Prediction.
• Decision Support.

3
Goals

• Given input conditions in a Heat Treating
  experiment,
• estimate the resulting Heat Transfer Curve.

2. Given the desired Heat Transfer Curve in an
experiment, estimate the conditions that
would obtain it.
4
Applications

• Estimated Heat Transfer Curve can be used to
• Get boundary conditions useful for simulations
  with tools such as DANTE.
• Obtain the corresponding time-temperature curve
  which when superimposed over a Jominy end quench
  graph predicts microstructure.
• Estimated input conditions can be used to
• Select process parameters and products for Heat
  Treating in the industry.
• Hence, this computational estimation enhances
  decision support in Heat Treating.

5
State-of-the-art

• Mathematical Modeling M-95, S-60, PG-60
• Variables not known in existing models in some
  cases.
• Precise equations not known in some cases.
• Similarity Searching HK-01 WF-00 M-97
• Naïve Search Non-matching conditions could be
  significant.
• Weighted Search Precise weights denoting
  relative importance not known apriori.
• Case-based Reasoning AP-03, K-88, AV-01,
  S-88.
• Regular CBR Involves human intervention.
• Exemplar Reasoning Problem similar to naïve
  search.
• Instance-based Reasoning Problem similar to
  weighted search.

6
Proposed Approach AutoDomainMine

• Performs domain-type-dependent data mining to
  discover knowledge for estimation.
• Learning analogous to domain scientists.
• Steps
• Knowledge Discovery Grouping graphs, reasoning
  causes of similarity to build representative
  cases.
• Estimation Use the reasoning and representative
  cases to estimate new cases.

7
Learning Analogous to Scientists
Grouping
Reasoning
Cause of Similarity
Input Conditions
Source CHTE
8
Step 1 Knowledge Discovery
9
Step 2 Estimation
10
Clustering

• Clustering Technique that places a set of
  objects into groups of similar objects HK-01.
• Used to group Heat Transfer Curves as a domain
  scientist would.
• After clustering, labels become available as
  Cluster IDs.

.
Clustering
11
Classification

• Classification is a technique to extract models
  to predict categorical labels HK-01.
• After clustering, labels available, so
  classification can be applied.
• Decision tree classifiers used in AutoDomainMine.

Sample Partial Output of Clustering
Decision Tree Classifier
Snapshot of Partial Decision Tree created
12
Classification (Contd.)

• Decision Trees form basis for selecting
  representative cases.
• Process
• From all paths to a cluster, select any one as
  representative conditions.
• From all graphs in the cluster, select any one as
  a representative graph.
• Decision Trees and representative cases used to
  estimate new cases.

Sample Partial Decision Tree
Sample Representative Case
13
Estimation Example 1
14
(No Transcript)
15
Estimation Example 2
16
(No Transcript)
17
Summary

• AutoDomainMine proposed as a technique for
  computational estimation of Heat Transfer Curves.
• This aids microstructure prediction and decision
  support.
• Ongoing Work
• Learning a domain-specific distance metric for
  clustering to improve accuracy.
• Designing domain-specific representative cases
  for better classification.