10-04-2025, 11:59 PM
This is a great ambitious project. From an engineering standpoint, you're fighting several battles:
Tolerance & Surface Finish: FDM printing inherently has layer lines that act as leak paths. Your sealing surfaces need to be smooth. Consider designing your valve for a press-fit silicone diaphragm instead of a ball. The servo would actuate a plunger that presses the diaphragm onto a seat. This is a much more reliable sealing method for additive manufacturing.
Creep: Both PLA and PETG will creep (deform over time) under the constant pressure of the seal, eventually leading to failure. This is another point for Nylon or PP (Polypropylene), though both are challenging to print.
Stall Torque: Calculate the required torque. The SG90 has ~1.8 kg-cm. With a 4mm orifice and even low pressure (e.g., 10 PSI), the force on the ball can be enough to stall it. Your stalling issue is almost certainly a combination of friction and fluid pressure.
I second the O-ring suggestion. Redesign for a seal; it's non-negotiable for a reliable valve.
Tolerance & Surface Finish: FDM printing inherently has layer lines that act as leak paths. Your sealing surfaces need to be smooth. Consider designing your valve for a press-fit silicone diaphragm instead of a ball. The servo would actuate a plunger that presses the diaphragm onto a seat. This is a much more reliable sealing method for additive manufacturing.
Creep: Both PLA and PETG will creep (deform over time) under the constant pressure of the seal, eventually leading to failure. This is another point for Nylon or PP (Polypropylene), though both are challenging to print.
Stall Torque: Calculate the required torque. The SG90 has ~1.8 kg-cm. With a 4mm orifice and even low pressure (e.g., 10 PSI), the force on the ball can be enough to stall it. Your stalling issue is almost certainly a combination of friction and fluid pressure.
I second the O-ring suggestion. Redesign for a seal; it's non-negotiable for a reliable valve.