INTRODUCTION

A wide variety of process conditions and parameters influence the design of high-performance gas phase contamination control solutions such as filters to remove airborne molecular contamination (AMC). AMC filters play a pivotal role in removing harmful contaminants from air streams in commercial operations, to protect humans, products and processes. Proper filter pleat designs maximize lifetime of the filter and provide lower costs of ownership. However, it is difficult to obtain filter performance experimentally at very low contaminant concentrations due to long test times, high experimental cost and technical limitations.

Virtual prototyping using computational fluid dynamics (CFD) can minimize these limitations. The link between AMC filter performance and a predictive, numerical model is validation through testing. A numerical model is only as good as its verified performance. Physical measurements are essential to provide a sound basis, to which the model can be tuned, before it is able to extrapolate the physical measurements to calculated data.

This study will demonstrate an approach that involves validation and prediction of the removal efficiencies of AMC filters at low contaminant levels required in many industrial applications such as SEMI cleanrooms. The application introduces ...

MATERIALS AND METHODS:

In our test setups, the AMC filter is positioned in a laminar air flow zone inside a wind tunnel. The pollutant test gas is introduced in the upstream airflow. The filter is then challenged with the test gas at controlled concentrations, flow rates & test conditions. The test gas is measured in both upstream and downstream locations to measure the filter performance. Fig. 1a is a sketch of the test set up, Fig. 1b shows the model of a single layer pleated filter.

Activated carbon was chosen as the adsorbent and toluene was chosen as proxy organic contaminant with initial upstream concentration of 1 ppm. A 2D representative transient CFD model was developed with a goal to predict AMC removal at low parts per billion (10^-9, ppb) level, as the ability to measure AMC at those test levels is important to understand behavior in ...

RESULTS & VALIDATION

An unsteady 2D CFD pleated filter model was developed by ANSYS Fluent for a specific air flow rate with toluene as the contaminant. The flow was laminar and the model was based on gas adsorption processes in a porous adsorbent bed. Modeling pressure compared well with experimental values. Moreover, the flow profile offered spatial and temporal resolution of the flow which is critical in assessing local flow uniformity profiles of filter systems. The models were able to predict and validate filter adsorption capacity. Figures 2 captured the transient contaminant mass fraction progress at three different time stamps (5, 111 and 236 hours), respectively. Red and blue indicate the saturated and unused parts of the media...

CONCLUSIONS

The design of high-performance gas filtration and purification products critically rely on the sensitivity of adsorption performance to a variety of process parameters, which are often difficult to know in advance of product design. Due to the constraints of conducting experiments, a CFD methodology was developed to predict adsorption behavior and filter capacity of specific AMC pleated filters. Results indicated that ...