DARREN FORSCHINO
Home / Case 01 — Northrop Grumman
Northrop Grumman · Engineering Experience

Mechanical &
Thermal Engineering.

Developed and iterated conduction-cooled electronics packaging through first-principles analysis, CAD, thermal FEA, design reviews, manufacturing support, and physical fit-up.

The underlying program is proprietary. This case study is organized around the engineering process rather than the product: what the problem was, how I worked it, and what I took from it.

Disclosure
Technical details and visuals are intentionally generalized to protect proprietary information. Hardware geometry, requirement values, component placement, temperatures, and internal imagery are omitted.
Role
Mechanical Engineering Intern · Mechanical / Thermal Focus
Focus
Electronics packaging
Tools
Siemens NX · Simcenter
Disclosure
Sanitized
FIG. 01Engineering loop
01Requirements & constraints
02First-order thermal model
03CAD concept / variants
04Mesh + boundary conditions
05Steady-state thermal FEA
06Interpret results / bottlenecks
07Technical design review
08Design iteration
09Drawings / mfg support
10Physical fit-up
11Issue correction / revalidate
Sanitized workflow — created for this portfolio.
§ 01 — Engineering Problem
Coupled constraints

A circuit-card assembly needed mechanical geometry that also carried heat.

I was assigned an electronics-packaging design problem: a circuit-card assembly installed inside a constrained chassis architecture. The mechanical solution had to protect the electronics, locate and retain the assembly, hold its interfaces, and provide a conductive heat path from heat-generating components — all while staying manufacturable and mass-efficient, and satisfying thermal requirements under a worst-case hot, steady-state condition.

Thermal performance, packaging, interfaces, manufacturability, and mass were coupled. Changing wall thickness, contact geometry, material distribution, or interface features improved one objective and generally degraded another.

FIG. 02Coupled constraints
THERMAL
PACKAGING
INTERFACES
MANUFACTURING
MASS
Generic constraint categories. No proprietary values shown.
§ 02 — My Contribution
Individual ownership
01

Mechanical Design

Developed CAD geometry in Siemens NX. Iterated wall thicknesses, contact features, material distribution, interfaces, and packaging against thermal and manufacturing constraints. Produced manufacturing drawings.

02

First-Principles Analysis

Built simplified 1-D thermal-resistance models in hand calcs and Excel. Used them as first-order design guidance and as a sanity check against simulation trends.

03

Thermal FEA

Developed and executed the Simcenter thermal-analysis workflow, covering geometry prep, meshing, boundary conditions, and worst-case hot steady-state solves. Used the results to drive design variants.

04

Design Reviews

Presented alternatives, assumptions, simulation results, and tradeoffs in Technical Design Reviews. Incorporated feedback from more experienced engineers into subsequent iterations.

05

Manufacturing Support

Updated geometry to improve manufacturability, supported drawing development, and coordinated with fabrication as design intent translated into hardware.

06

Hardware Integration

Participated in physical assembly and form/fit evaluation. Applied thermal interface materials during integration; diagnosed and corrected fit interferences on early prototype units.

§ 03 — First-Principles Thermal Analysis
Before FEA

Establishing intuition before detailed simulation.

Before relying on detailed FEA, I used simplified one-dimensional conduction models to understand the dominant thermal behavior: approximate resistance through candidate conduction paths, sensitivity to wall thickness, material distribution, and whether a proposed geometry was moving performance in the correct direction.

Hand calculations weren't a replacement for FEA. They were the fast engineering check that built intuition and screened geometry changes before evaluating detailed variants in simulation.

FIG. 031-D thermal resistance
HEAT SOURCE — component
Rinterface
THERMAL INTERFACE — TIM
Rcover
METALLIC COVER — conduction path
Rchassis
CHASSIS BOUNDARY — sink
Generic resistance chain — illustrative only. Not reconstructed from proprietary geometry.
§ 04 — Thermal FEA in Simcenter
Worst-case hot steady state

Developed the Simcenter thermal-analysis workflow.

Prepared geometry, developed and refined meshes, applied documented boundary conditions, evaluated worst-case steady-state conditions, interpreted temperature distributions and thermal bottlenecks, and used the results to drive mechanical design iterations.

A significant part of the work was model discipline: simplifying and meshing efficiently without inflating computational cost, and reading temperature distributions for what they revealed about the geometry rather than only the peak value.

FIG. 04Sanitized FEA workflow
01
Prepare geometry
02
Mesh & refine
03
Boundary conditions
04
Conductive interfaces
05
Solve
06
Read distributions
07
Identify bottlenecks
08
Compare variants
09
Review → update mechanical design → loop
.
§ 05 — Designing the Conduction Path
Mechanical geometry as heat path

Establishing a Lower-Resistance Conduction Path.

One important design strategy was creating direct conductive paths between higher-heat components and the metallic cover through thermal interface material, a lower-resistance route than relying primarily on heat spreading through the circuit-card structure.

I iterated the contact geometry and surrounding material distribution against thermal results and packaging constraints. Once modeled performance was adequate, material that contributed little to the dominant thermal path could be reduced, cutting mass and manufacturing complexity without giving up the intended conduction behavior.

The mechanical design and the thermal design were developed together rather than sequentially, because the conduction path was defined by the geometry.
FIG. 05Conduction path concept
COMPONENT
TIM
METALLIC COVER
CHASSIS
heat source
interface
conduction path
boundary
Q heat out
Generic conceptual path. Illustrative only.
§ 06 — Design Iteration & Tradeoffs
Every change moved multiple axes

The design evolved through repeated analysis and review. Variables in play included wall thickness, material distribution, contact-feature geometry, mechanical interfaces, cover-to-cover interface concepts, retention configuration, thermal-interface implementation, manufacturability, and packaging clearance.

The important point isn't the specific configuration that was selected — it's the coupling:

Each mechanical change altered the thermal network, and each thermal improvement had to remain geometrically compatible and manufacturable.
FIG. 06Coupled trade
THERMAL
performance
VS.
MASS
PACKAGING
MANUFACTURABILITY
§ 07 — Technical Design Reviews
Analyze → review → iterate

I presented proposed geometry, analysis assumptions, simulation results, and design alternatives in formal Technical Design Reviews — comparing configurations, discussing where observed behavior differed from expectation, taking feedback from more experienced engineers, and updating assumptions, models, or CAD in the next iteration.

I drove the analysis and design work under engineering supervision; the review process shaped which tradeoffs were carried forward.

FIG. 07Review loop
ANALYZE
REVIEW
ITERATE
loop
§ 08 — Design for Manufacturing
Easy in CAD ≠ good geometry

Geometry that is easy to create in CAD is not automatically good manufacturing geometry.

During iteration I learned to identify and remove unnecessary manufacturing complexity: features smaller or more intricate than needed, internal geometry that created unnecessary machining difficulty, and tolerances or shapes whose engineering value didn't justify the fabrication cost.

The design was revised to preserve mechanical interfaces and thermal function while simplifying noncritical geometry.

RETAINED
  • Mechanical interfaces
  • Thermal contact geometry
  • Retention features
SIMPLIFIED
  • Over-specified noncritical features
  • Internal geometry creating machining difficulty
  • Tolerances not carrying engineering value
§ 09 — From CAD to Hardware
Model verification via fit-up

A digital model doesn't guarantee physical fit.

During physical fit-up of an early manufactured unit, I found interference between the cover geometry and components on the circuit-card assembly. Physical fit-up exposed a discrepancy between the analysis geometry and as-built hardware.

01
Identified the interference during fit-up.
02
Referenced updated geometry and verified the revised model against the physical assembly.
03
Updated the cover geometry and rechecked form and fit.
04
Reassessed the design after the geometry change; the corrected design subsequently achieved the required fit.
Lesson
A simulation is only as trustworthy as its inputs.
Fit-up exposed a discrepancy that wasn't visible in the digital model, a concrete reason to validate analysis geometry against real hardware.
FIG. 08Model ↔ hardware loop
DIGITAL MODEL
BUILD
PHYSICAL FIT-UP
INTERFERENCE FOUND
MODEL CORRECTION
REVALIDATE
§ 10 — Outcome
Sanitized

The design progressed through analysis, review, manufacturing, and physical integration, with hardware used to validate form, fit, and the analysis-driven design approach.

§ 11 — Engineering Lessons
Five takeaways
01

Models are only as good as their inputs.

A high-quality analysis can still produce misleading confidence if the underlying geometry or assumptions don't represent the real hardware.

02

First-order analysis and FEA reinforce each other.

Hand calcs gave intuition and trend checks; simulation gave the spatial resolution to evaluate realistic geometry. Neither was used blindly.

03

Thermal and mechanical design are coupled.

Wall thickness, contact geometry, material removal, packaging clearance, retention, and manufacturability all moved the thermal solution.

04

Manufacturability is part of the design.

A geometrically valid CAD model isn't automatically a good physical design. Fabrication constraints changed how features were sized and shaped.

05

Iteration is normal engineering.

Design reviews, updated constraints, simulation results, manufacturing feedback, and hardware fit-up all caused legitimate design changes. The process was not linear.

Next case study
Journey to 10,000 ft