Thermal analysis of cryogenically assisted abrasive jet micromachining of PDMS

In cryogenic abrasive jet micromachining (CAJM), it is important to sufficiently cool the machined surface. However, over-cooling or cooling a larger volume than necessary, reduces the cost-effectiveness of the process. With a view to process optimization, a finite element thermal analysis has been made of the CAJM of channels and holes in polydimethylsiloxane (PDMS). For the machining of holes, using a previously developed experimental CAJM setup, it was found that the thermal front moved nearly 12 times faster than the material removal front, implying that the target PDMS was cooled sufficiently prior to micromachining. The time required to cool a chip removed by a single particle was calculated to be 0.7 ms while the time required to machine it was 7.7 ms indicating the possibility of optimizing the cooling process by reducing the heat flux. A finite element analysis (FEA) was used to simulate a moving heat sink to investigate the general cooling behavior during the machining of unmasked and masked channels and to assess the influence of the jet scanning speed. The temperatures predicted by the FEA agreed reasonably well with those that were measured using embedded thermocouples. Under steady-state conditions, the temperature distribution within the moving jet footprint was highly asymmetric, being much colder toward the trailing edge and the target temperature at a given position decreased with decreasing scan speed. A coupled thermal/structural finite element analysis was used to examine the influence of thermally induced strains on the final room-temperature shape of micromachined features. It was found that accounting for these thermal strains in the surface evolution model significantly improved the agreement of predicted with measured profiles.

► A thermal analysis of cryogenic abrasive jet micromachining was performed.
► There was good agreement between predicted and measured surface temperatures.
► The thermal front advanced faster than the material removal front for the machining of holes.
► For channel machining, the target temperature decreased with decreasing scan speed.
► The prediction of channel shapes improved when thermal distortion was modeled.