I had the chance to discuss fluid dispensing recently with my fellow-Brit, Steve Adamson of Asymtek.
[Andy C Mackie] What are the "natural limits" for dispensing fluids as sizes shrink?
[Adamson, Steve] Really, there are none. As dimensions shrink, we see formulators make changes to the fluids to help them flow. Underfills are a good example: several times we have seen people make pronouncements that you cannot use capillary underfills below a certain gap, only to see a new fluid that appears to work fine. I do think there is some limit but we will probably move to pre-applied at that time.
If you are talking about dot dispensing, there is a technology gap. Inkjet print guys with micro-fine inks can make small dots, but they cannot use fluids with greater than 20cps viscosity (the viscosity is mainly driven by solids loading). This limits the current-carrying capability for metal-filled fluids and pastes. Dispensing can take fluids with much greater viscosity and dispense dots or lines. The diameter limit for small dispensed dots is around 125 microns. But this is highly dependent on fluid rheology and many other factors.
[Andy C Mackie] What do you see as the primary differences between jetting and dispensing?
[Adamson, Steve] Dispensing through needles is a general work-horse technology. It's relatively inexpensive and the technology is well known. Jetting is faster because you do not have to have an accurate substrate to needle tip gap each time you dispense a dot of fluid. In the past we had to height sense every time a precision dispense was required. That is why many applications have moved to jetting.
Jetting delivers fluid to the substrate in a different manner than needle dispensing. In jetting you have to provide energy to get the dot to break away from the nozzle tip. When the dot hits the substrate it flattens out. When dispensing small dots with a needle you touch the end of a wetted needle to a substrate and break in half the fluid that is on the end of the needle. Dot sizes from jets tend to be larger than needles, but there are fewer complications from stringing, and there are no speed reductions from the extra z-axis motion to snap apart the fluid.
[Andy C Mackie] When does jetting become the primary material deposition process?
[Adamson, Steve] Speed and tight dispense geometries are the two drivers here. Let me explain: it's sometimes difficult to get a needle between two devices you want to underfill. When a dot is jetted from a nozzle tip, there is no physical needle to be concerned about. It's a pure dot of fluid flying through the air. Typically this is about 100 microns in diameter, but can be smaller. The dot can be shot through the small gap. Then it hits the board and spreads out.
[Andy C Mackie] How does the odd rheology of many dispensed fluids (viscoelasticity / plasticity etc) play into the use of the Weber number, which just refers to "viscosity" with the assumption that the fluid is Newtonian?
[Adamson, Steve] This is a fluid expert question, and that is not my background, a long time ago I was an electrical engineer. We do know that some fluids jet better than others. We have run tests on various fluids, and have found that you can pound on some fluids and they do not change properties. Other fluids are more thixotropic and these generally appear to have better performance for jetting. As you have pointed out there are a lot of variables here and they all impact the end results.
[Andy C Mackie] Steve, I really appreciate your complete and very interesting response. Thank you.