Research Experiences For Undergraduates (REU)

REU 1999 Projects

Pricing of a Child Rider

Sponsor: John Hancock
Advisor: Arthur Heinricher and Anne Wiedie


Sarah Winne		Bryan Heilman		Christopher Shane

Currently, John Hancock's policy holders have the option of paying an additional six dollars per thousand dollars of coverage in order to add a child rider onto their insurance policy. In the case of a child's death, John Hancock pays the primary insured a death benefit. In the case of the death of the primary insured, John Hancock will then continue paying the annual fee of six dollars per thousand for the child until the child reaches age 21. The children also have the option of converting the policy to up to five times its original value without being underwritten once the children turn 21. This cost of six dollars per thousand dollars coverage remains constant regardless of the number of children in the household or the age of the children (after the children are older than 14 days).

However, this price of six dollars per thousand dollars of coverage was calculated thirty years ago using only very rough estimates. Thus, the goal of this project is to develop a method by which to accurately calculate the cost of this policy.

Optimal Shape Design in Metal Processing

Sponsor: Morgan Construction
Advisor: Bogdan Vernescu


Catherine Lee		Brendan Hogan

In most rolling mills, the metal rods are coiled by pushing them through a pipe that has a corkscrew-like shape, called the laying pipe. The goal of this project is to model the wear distribution along a laying pipe and to find the shape of the pipe that has the least wear. A lower and better distributed wear extends the life of the pipe and reduces the down time of the mill, thus increasing the productivity.

Given specific boundary conditions on the geometry of the pipe, we derived a shape dependent expression for wear from the forces involved. In finding the optimal shape that reduces the wear, we plan on using techniques such as calculus of variations, optimal control, and numerical methods to solve the differential equations involved.

Statistical Models Predicting Demand for Loans Against Permanent Life Insurance Policies

Sponsor: John Hancock Insurance
Advisor: Ann Wiedie and Art Heinricher


Xiomara Carrero		Samuel Pruitt		Lauren Wheeler

The purpose of this project is to find a model or several models that will allow John Hancock to understand and predict the loan demand policyholders of permanent life insurance policies will borrow against their type of policy. We will determine if there is a seasonal trend in demand for loans against John Hancock�s life insurance policies and we will identify the significant predictors of changes in loan demand. We will also determine if there is a time lag between a change in the predictor variables and a change in the demand for loans. The amount of loan assets is constantly changing, therefore, we need to have a dynamic model for this problem.

In order to validate the model, we will use data from 1970 to 1997 to predict 1998 loan demand. We will then check our error to see how close the predicted loan demand is to the actual loan demand.

Pressurized Pipe Modeling

Sponsor: Veeder-Root , Simsbury, CT
Advisors: Roger Lui


Brian Ball		Paul Dostert		Jake Woods

Undeerground pressurized pipe leak detecting can be done when a closed pipe is idle by using a single pressure sensor to observe the pressure decay caused by a leak. But pressure changes caused by inital temperature conditions driving thermal flow can either mask a leak or cause false alarms. The project proposes to develop mathematical models to accurately described thermal and leak induced pressure versus time data so that leaks can be distinguished from thermal interference.

Maintained by webmaster@wpi.edu
Last modified: Jun 20, 2010, 19:34 UTC