Thermal Analysis Of A Continuous Flow Polymerase Chain Reaction (Cpcr} Microdevice Using Finite Element Method And Neuro-Genetic
Loading...
Date
2006-03
Authors
Hing Wah, Lee
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
A successful polymerase chain reaction (PCR) process used for amplification of the
deoxyribonucleic acid (DNA) can only be achieved when the necessary temperature
and residence time requirement are fulfi"ed. In view of this, temperature analysis and
subsequently temperature control must be performed. Recent research focused more
on the fabrication and the development on CPCR microdevice while neglecting
thorough analysis on the temperature characteristics. In addition, explanation and
presentation of the temperature control methods are almost non-existent.
To eliminate the above limitations, a three-dimensional finite element analysis
(FEA) and a two-dimensional analytical finite element model has been developed to
analyze the thermal performance of the CPCR microdevice. For the FEA, two types of
modeling, namely the detailed and the simplified model, have been used while only the
simplified model is used for the analytical simulation. Comparison of results between
temperatures obtained using the ANSYS® FEA (detailed and simplified model) and the
analytical FEM simulation (simplified model) with the published experimental results
showed good agreement. However, simulations using the simplified model predicted a
lower temperature than that of the detailed model due to the higher convective heat
transfer area present in thesimplifiedmodel'and the effect of the DNA buffer flow
separation in the detailed model. From the temperature analysis, it can be inferred that
to obtain a successful PCR process for a constant flow velocity of 1.3mm/s, the input
power must be assigned with a value between 560mW - 580mW, 60mW - 80mW and
300mW - 31 OmW for heater 1, heater 2 and heater 3 respectively.
This work also introduced a novel temperature control methodology using
Neuro-Genetic optimization which combined the artificial neural network (ANN) and
genetic algorithms (GA). Results for temperature predictions obtained using ANN
showed good agreements with the finite element simulated results. From the GA
analysis, the optimized result with the corresponding vaiue of 580mW, 80mW and
306mW for heater 1, heater 2 and heater 3 respectively is obtained where it is abiG tel
fulfil both the temperature and residence time requirement.
Description
Keywords
Engineering , Machanical