Cleaning Fixture Design: Datum, Part Protection And Commonality

Overview

A cleaning fixture is the least visible part of a washing system and the one most likely to be designed last. It holds the part, it decides where the fluid goes, and it decides whether the part is damaged before it is clean. When a line underperforms, the pump is checked first and the fixture is checked last, which is usually the wrong order.

Big Bird Industrial (formerly Shimada Big Bird Industrial) has built cleaning fixtures for powertrain components since 2001. Three principles account for most of the difference between a fixture that works and one that is quietly responsible for scrap: datum that matches the drawing, contact points that protect the part, and commonality that survives a model change.

The fixture is also a flow device. Every surface it covers is a surface that does not get cleaned, and every pocket it creates is a place where fluid stays. Design for cleaning and locating at the same time, and the two stop fighting.

1. Datum: Locate The Way The Part Is Machined

A cleaning fixture does not need to measure the part, but it does need to position it repeatably, because the nozzles are aimed at specific features. If the part sits differently on each cycle, a targeted jet that was aimed at an oil gallery will miss it on one part in ten.

The reliable datum is the one the part was machined from: two locating holes and a face, or three points on a machined surface. Locating on a raw casting surface is the common shortcut and the common failure, because casting draft and parting-line flash vary by more than the position tolerance of the features being cleaned.

A second consideration is over-constraint. A fixture that locates on six surfaces will fight itself when the part is at the upper end of its tolerance, and the part will sit on a diagonal rather than flat. Three-two-one is still the rule that works.

2. Protection: Contact Points Do Damage

Every contact point is a potential witness mark, and on a machined or painted surface a witness mark is a defect. The parts of a fixture that touch the workpiece should be chosen for softness and for replaceability.

Materials that work: engineering plastics for light parts, urethane pads for painted surfaces, and hardened steel only where the surface is non-critical. Materials that cause trouble: bare steel on aluminium (galling), rubber on hot parts (degradation and transfer), and anything that holds water against the surface.

Contact points also trap fluid. A pad that sits flush against a surface leaves a film that the rinse cannot reach, and that film dries into a residue spot. Grooved or dimpled pads drain; smooth pads do not.

3. The Fixture As A Fluid Device

A fixture that blocks a jet does not clean the surface behind it. Before the nozzle layout is fixed, the fixture should be reviewed for shadowing: which surfaces are hidden from every nozzle by a clamp, an arm or a plate.

The practical technique is to mark up a section drawing with the fixture in place and trace each jet. Surfaces that receive no direct spray must be reached by turbulence or by a secondary nozzle, and that decision has to be made on paper, not after commissioning.

Drainage follows the same logic. The fixture should present the part in the attitude that lets fluid leave, or provide a rotation that does. A fixture that holds a pocket level will fill it, and the residue will appear exactly where the fixture was.

4. Commonality: Design For The Next Part

Cleaning lines outlive the parts they were bought for. A fixture designed for one variant becomes a constraint the first time a new variant arrives, and the cost of that constraint is paid in lost flexibility rather than in capital.

Three approaches, in increasing order of cost: locate on features shared across the family; build interchangeable nests on a common base; or accept a dedicated fixture per variant. The first is best and is often possible because a family shares bearing bores or a deck face.

Changeover time matters as much as the mechanism. A nest that takes four hours to swap will not be swapped, and the line will run one variant while the second is cleaned elsewhere. Quick-change features that pay for themselves: captive fasteners, dowel location with no adjustment, and utilities that connect without tools.

5. Verifying A Fixture

A fixture is a component and it needs its own acceptance, separate from the washer's. The checks are simple and frequently skipped.

The dyed fluid test is the most useful of the four and the least used. A small amount of tracer in the bath, followed by inspection under UV, shows exactly which surfaces were reached and which were shadowed, and it takes an hour.

6. Maintenance And Wear

Fixtures wear, and they wear in ways that change the process. Plastic pads compress, locating pins wear oval, and clamps lose force. None of these produces a visible failure; they produce a gradual drift in position that shows up as a cleaning defect in one region and not another.

A wear list with check intervals is the practical countermeasure: pads quarterly, pins every six months, clamp force twice a year. The alternative is to discover wear when a new variant will not fit, which is a much more expensive way to find out.

7. When To Automate The Fixture

Manual loading is fine while the part is light and the volume is low. Above a certain mass, or above a certain takt, the fixture becomes an automation project: clamping, rotation and transfer all need actuators, and the design constraints tighten because an automated clamp can crush a part.

The transition point is usually where the operator becomes the cycle-time constraint rather than the machine. At that point the fixture design question changes from how to hold the part to how to present it to a robot, and the locating scheme has to survive being approached from a taught position rather than a human hand.

8. Documentation That Outlives The Project

The fixture drawings that matter after commissioning are not the assembly drawings. They are the contact-point list, the clamp force specification and the wear-part schedule, because those are what maintenance needs and what a new variant has to be checked against.

A contact-point list also settles arguments. When a witness mark appears on a finished part, the question is whether the fixture touched that surface, and a list answers it in a minute rather than in a meeting.

Finally, the fixture should carry its identity: a plate with the part family, the revision and the changeover instructions. Fixtures that lose their identity get modified ad hoc, and an ad hoc modification is how a validated process becomes an unvalidated one.

Frequently Asked Questions

Should the fixture locate on machined or cast surfaces?

Machined, wherever possible. Cast surfaces vary by more than the position tolerance of the features being cleaned, so locating on them makes the process unrepeatable.

How much clamping force is enough?

Enough to hold the part against jet reaction without marking it. That is a measurable requirement, and it should be stated as a range rather than left to the fitter.

Can one fixture serve a whole part family?

Often yes if the family shares locating features. Where it does not, interchangeable nests on a common base give most of the flexibility for a fraction of the cost of a second machine.

Why do parts come out with clean faces and dirty pockets?

Because the fixture shadows the pockets or holds them level so they cannot drain. Both are visible on a drawing before the fixture is built.

Conclusion

A cleaning fixture is a locating device, a protection device and a fluid device at the same time. Datum decides repeatability, contact points decide damage, geometry decides whether the fluid reaches the surface, and commonality decides whether the line still works when the part changes.

Design it with the nozzle layout rather than after it, verify it with tracer fluid rather than with a visual check, and put it on a wear schedule. The rest of the washer will then perform the way it did on acceptance day.