Skip to content

Project 02 — January–May 2026

Mitsubishi 6G72 V6 — CAD Assembly

A fully mated, motion-validated V6 SOHC engine reverse-engineered in SolidWorks from part drawings and reference dimensions — crankshaft, camshaft, valve train, and a 20+ part accessory drive, modeled and assembled from scratch.

Cutaway view of the V6 assembly showing pistons, crankshaft, timing belt and both camshafts

Cutaway — pistons, crank, timing belt, both cam banks

Unique part types
40+
Cylinders, V-configuration
6
Bore
91.1mm
Stroke
76mm
SolidWorksMotion StudyGD&T / 2D DrawingsReverse Engineering
01

Why this project

I wanted deliberate practice at two things a robot-arm-scale project doesn't exercise: reading real part drawings and reference dimensions instead of designing from a blank sketch, and managing a large assembly where a single mistake in one subsystem (a bore diameter, a journal offset) breaks how a dozen other parts mate.

The 6G72 was the target: a real Mitsubishi V6 SOHC engine, reverse-engineered feature by feature — cylinder block, crankshaft and connecting rods, camshafts and the full valve train (rocker arms, followers, springs), timing belt and cam gears, intake and exhaust manifolds. Every dimension was pulled from reference measurements and constrained with real geometric mates, not eyeballed.

Mechanical motion was validated with a SolidWorks Motion Study — the crankshaft actually drives the pistons and valve train through the modeled constraints, not just a static exploded view. Tolerancing wasn't applied given the project timeline; the emphasis here is assembly complexity and correct kinematic relationships between parts.

02

How the power flows

The assembly isn't a static shell — every rotating part is constrained to the one before it. Step through the chain to see what drives what.

Combustion drives the pistonsThe crankshaft turns the belt pulleyThe belt synchronizes both cam banksCams push rockers, rockers open valves

Step 1 of 4

Combustion drives the pistons

Each piston travels up and down in its bore, with its connecting rod constrained concentric to the crankshaft journal. Six pistons across two banks put rotational force into the crank.

03

Model views

The completed assembly from three angles — intake and exhaust manifolds, timing belt, and both cylinder banks in place.

Isometric view of the V6 assembly

Isometric

Front elevation view of the V6 assembly

Front elevation

Rear isometric view of the V6 assembly

Rear isometric

04

The assembly

SolidWorks screenshots of the actual model, plus a dimensioned 2D drawing and the full exploded-view BOM. The reference photo of the real 6G72 is included for comparison, not as my own work.

05

Motion study

Crankshaft rotation drives the pistons and, through the timing belt and cam gears, the full valve train — validated as an actual kinematic simulation, not a rendered animation.

SolidWorks Motion Study — slowed to 0.75× for clarity

06

Key dimensions

Cylinder bore91.1 mm
Cylinder stroke76 mm
Valve spring30–34 mm
Rod journal50 mm
Main journal60 mm
ConfigurationV6, SOHC
07

Bill of materials

A partial view of the part list — enough to show the scale of what has to mate correctly for the assembly to move as one system.

PartQty
Engine block1
Crankshaft1
Crank bushing2
Camshaft1
Camshaft (opposite side)1
Valve12
Valve washer12
Rocker arm12
Rocker follower12
Cylinder head2
Cylinder head cover2
Piston (part 1–4 sets)24
Lower block1
Oil pan1
Intake manifold1
Exhaust manifold + mirrored2
Exhaust manifold gasket2
Head gasket2
Cam gear2
Flywheel1
Timing belt1
Pulley2
Air filter1
08

What I'd build next

What this model proves is synchronized mechanical motion — multiple axes of rotation driven from a single input, with pulleys and gears enforcing the timing relationships between them.

The obvious next step is the part I deliberately left out: the combustion system. Modeling the fuel path, injectors, and spark timing would turn this from a kinematic assembly into something that could actually be simulated as a running engine — which is a meaningfully harder problem than geometry, and the one I'd want to take on next.