When should DFMA merge several machined parts into one, and when does part consolidation backfire?

How do you know an assembly of machined parts is a part consolidation candidate?

The usual signs are an assembly whose labor, fastener and inspection cost rivals the cost of its parts, and a critical alignment that drifts because it depends on many interfaces in series. If the bill of materials is mostly screws, spacers and small brackets that never move, the design is probably carrying parts it does not need.

Look for these patterns on the bill of materials and the shop floor. A bracket is bolted to a plate that is bolted to a housing, and all three are aluminum and never separate in service. Shims appear on the assembly instruction, or the line adjusts parts until a gap looks right. A critical bore, pin or face is located through two or three bolted joints instead of being machined relative to its datum. Fastener counts grow with every engineering change, and each new screw brings a tapped hole, a torque value, a thread check and a part number to stock.

None of these proves the parts should merge. They prove the assembly deserves a part-by-part review, which is exactly what the DFMA method was built to do. Boothroyd Dewhurst describes the DFA Index as a comparison between the theoretical minimum assembly effort and the actual effort of a design (Boothroyd Dewhurst, design for assembly), so a low score points at parts and operations worth questioning before anyone redraws anything.

How do you decide whether two machined parts should become one?

Ask the three minimum-part questions for each part in turn: must it move relative to the parts it touches, must it be a different material or process, and must it stay separate so the product can be assembled, adjusted or serviced? A part that answers no to all three is a candidate to merge or eliminate, and a part that answers yes to any one has earned its place.

The three questions come from the Boothroyd Dewhurst method and are set out on the Boothroyd Dewhurst DFA page and in the textbook by Boothroyd, Dewhurst and Knight. Work through the assembly in its build order, starting from the base part, and record the answer and the reason for each part. The count of parts that survive is the theoretical minimum, and the gap between it and the real count is the opportunity.

A practical way to run the review on an existing assembly:

1. List every part, including each screw, washer, pin and shim, in build order, with its material and how it is held.
2. Answer the three questions for each part and write the reason next to it. Mark fasteners separately, since they are rarely needed as parts in their own right.
3. For each critical alignment, trace the chain of features from the datum to the controlled feature and count the joints and clearance fits it crosses. A chain that crosses two or more bolted joints is the first place to merge parts or add positive location.
4. Group the candidates into merges that share one material and one datum reference frame, then apply the four practical checks below.
5. Before redrawing, price the change. Boothroyd Dewhurst recommends pairing the assembly review with piece-part should-cost so each simplification is checked against what the new part costs to make.

Then run four practical checks that the three questions do not cover, because a part that is theoretically unnecessary can still be expensive to merge:

• Can it be made in one piece? Check that every feature of the merged part can be reached by a cutter, ideally from a small number of setups. A merged part that needs a long-reach tool into a deep cavity, or five setups instead of two, may cost more than the parts it replaces; the design for machining guide covers tool access and setup counting.
• What does the stock look like? Two thin parts merged into one thick block can mean machining most of a large billet away. Compare the bounding box of the merged part with the sum of the original blanks.
• Can it be inspected? Make sure the critical features remain reachable by a probe, gauge or optical measurement once the walls that used to be separate parts are joined.
• What does a scrap cost? One bad feature now scraps the whole part, so a merged part with an expensive feature next to a risky one raises the value of every rejection.

How do you cut fasteners and add self-locating features when parts must stay separate?

Let the geometry locate the part and use fasteners only to hold it. A pilot diameter, a pair of dowel pins or a machined step sets the position, and fewer, larger, standard screws carry the clamp load.

Bolt clearance is meant for assembly, not for location, so a part held only by screws in clearance holes can sit anywhere within that clearance. Replace that float with a deliberate locating scheme: one round pin and one diamond or slotted pin, a turned pilot that fits a bored register, or a shoulder that the part is pushed against. Dimension the locating features to the same datum reference frame as the features they control, using the datum and position language of ASME Y14.5; the press-fit and dowel pin guide covers the fits.

Then reduce the fastener count. Ask whether a joint carrying four screws actually needs four, whether two screw sizes could become one, and whether a captive nut, a tapped hole in the mating part or a retaining feature can replace a loose nut and washer. Standardize on one thread size and one drive type across the assembly so the line uses one tool. Add chamfers and lead-ins on pins and pilots so parts drop into place, and make asymmetric parts obviously asymmetric so they cannot be installed backward. Each of these lowers assembly effort without changing which parts exist.

What should the drawing and RFQ say for a consolidated part?

Send the merged part and the assembly it replaces, mark which features came from which original part, and dimension every critical feature from one datum reference frame. Tell the supplier what the consolidation was meant to achieve, so the quote can confirm or challenge it.

• Datums. One datum reference frame for the whole part, chosen from functional mounting features, with critical bores and faces controlled by position, perpendicularity or profile to it. The datum reference frame guide explains how to pick them.
• Critical-to-function features. Flag the features that used to depend on assembled joints, so they are measured as part of first article inspection rather than assumed.
• Material and condition. One material, named by alloy and temper, confirmed to suit every function the merged part now carries.
• Both options, if undecided. Ask for a price on the merged part and on the original parts, at the same volume, so the decision rests on quotes rather than assumptions.
• Service requirement. State which items must remain replaceable in the field, so a supplier suggestion does not merge them away.
• Process route, if casting is an option. Ask for the casting and finish machining to be quoted as one route with the datum scheme agreed between them.