Eduardo Cavazos Quintero

I'm Eddy, an aerospace engineering student at MIT. On MIT Motorsports I'm the responsible engineer for the rear wing, active aero, and sidepods, and on MIT's Sustainable Engine Team I'm the turbine team's responsible engineer. At the Wright Brothers Wind Tunnel, I'm implementing wall-interference and blockage corrections.

Résumé (PDF) See my work LinkedIn

Studying: B.S. Aerospace Engineering, MIT, 2029
GPA: 4.75 / 5.0
Looking for: Summer 2027 internships in aero, CFD, FEA, and propulsion
Languages: English and Spanish, native

Full-car CFD result for the MIT Motorsports MY26 car, with surface pressure coefficient colored from blue suction to red stagnation — MY26 car, surface pressure coefficient (STAR-CCM+)

Highlights

Rear wingRaised 2D rear-wing downforce 9% (328 → 358 N/m) for 3% more drag, lifting L/D from 16.5 to 17.4, by repositioning the five elements inside the existing endplate and comparing both layouts in 2D RANS CFD.
Radiator CFDShowed our stock radiator core rejects 2.89 kW against the 2.34 kW requirement, a 24% margin that justified an 8 mm thinner core saving 0.5 kg in total, by building a STAR-CCM+ model validated against the radiator's 93 Pa air-side and 5.1 kPa coolant-side pressure drops.
Gas turbineCaught 5 issues in our turbine's mean-line worksheet, including a stator hub exit Mach of 0.976 over the 0.95 limit (the sheet showed 0.952), by writing a MATLAB check that rebuilds the stage and matches its velocity triangles to within 0.01%.
Active aeroTook the MY27 active front wing through its concept review: 23 functional requirements, a breakdown of 60+ FSAE 2027 rules, a rod linkage chosen over belts and pulleys, and a budget of about $180 and 0.4 kg on the wing.
MY27 aero packageScoped a full aero redesign toward C_LA 4.0 ± 0.2 m², C_DA 1.3 ± 0.1 m², and a 42/58 front/rear balance within a 14.22 kg, $4,100 budget, by directing the aero team's design meetings and checking each change against the FSAE 2027 rules.
Wind tunnel researchImplementing wall-interference and blockage corrections in MATLAB and XFOIL at MIT's Wright Brothers Wind Tunnel, and leading my own airfoil study that compares predicted and measured lift, drag, and pressure across 10–15 angles of attack.

Selected work

Most of this is from MIT Motorsports, where our aero team is designing the MY27 car. When a number is a design target rather than a result, it's labeled as one.

MY27 aero architecture

MIT Motorsports, architecture design review, September 19, 2026

ArchDR deck (PDF, 45 slides)

My role: Directed the aero team's design meetings
Presented by: 6 members of the aero subteam
Review: Architecture design review (ArchDR)
Rules: FSAE 2027, T.7 aero and bodywork
Next gates: Concept reviews: sidepods, active aero

Downforce, C_LA: 4.0± 0.2 m²
Drag, C_DA: 1.3± 0.1 m²
Aero balance: 42 / 58front / rear
Yaw: ≤ 10%C_LA loss at 5°
Aero mass: 14.22 kg± 0.25
Budget: $4,100

MY27 functional requirements from the ArchDR. These are design targets, not test results.

Scoped a full aero redesign toward these targets by directing the aero team's design meetings and checking every proposed change against the FSAE 2027 aero rules.
Triaged 6 parts of the package by build effort with the team, with go/no-go gates where they're needed, so the package stays complete even if a high-risk item like active aero (capped at 0.3 kg of added mass) is dropped.

The package manages flow front to back across 7 elements. The front wing makes about 39% of the downforce and turns air around the front tires. Endplate sides and flaps push the tire wake outward, and the nose cone and whiskers guide clean air along the body. The sidepods recover pressure lost in the front-wheel wake and feed the radiators. Floor edge strakes keep tire wake out from under the car so a larger diffuser can make more suction, and the rear wing closes it out on a new gooseneck mount that clears the diffuser exit.

The review also responds to MY26's manufacturing lessons, from wet layups with the wrong amount of epoxy to mounts not designed for load and a surface finish rough enough to hurt CFD correlation.

MY26 baseline CFD

Scenes from the MY26 car's full-car STAR-CCM+ simulation, the baseline the MY27 targets build on. Pick a scene, and select the image to enlarge it.

Streamlines colored by velocity, with the car's surface colored by pressure coefficient. The disturbed lines trailing the car are its wake.

Velocity magnitude (m/s)< 515> 25

Surface pressure coefficient, C_p< −10> 1

Streamlines

Near-ground streamlines

Surface pressure

Total pressure

ArchDR cover slide — Cover, September 19, 2026

ArchDR slide listing the functional requirements: lift and drag area, aero balance, cost, and mass — Functional requirements

ArchDR slide on how the front wing, endplates, nose cone, and sidepods manage flow — Front-to-back flow management, part 1

ArchDR slide on floor edge strakes, the diffuser, and the rear wing — Front-to-back flow management, part 2

ArchDR slide triaging sidepods, the larger diffuser floor, and active aero by effort and go/no-go criteria — Effort and go/no-go triage

ArchDR slide summarizing the chosen architecture and the reasons for it — Architecture summary

MY27 rear wing element study

MIT Motorsports, rear wing, 2D CFD, 2026

My role: Rear wing responsible engineer; ran the CFD
Method: 2D RANS, k-ω SST, 15 m/s
Elements: 5: an S1223RTL main and four S1223 elements
Constraint: Fits inside the MY26 endplate
Status: Proposal for MY27

Downforce: 358N/m, from 328
Drag: 20.5N/m, from 19.9
Lift-to-drag: 17.4from 16.5
Trade: +9% / +3%downforce / drag

2D CFD results per meter of span at 15 m/s, MY27 proposal against MY26.

Raised 2D rear-wing downforce 9% (328 → 358 N/m) for 3% more drag, lifting L/D from 16.5 to 17.4, by repositioning the five elements inside the existing MY26 endplate and comparing both layouts in 2D RANS (k-ω SST) at 15 m/s.
Traced the gain element by element: the main element and first lower flap add about 17 and 11 N/m, nearly all of the 30 N/m gain, while the upper flap gives about 5 N/m back.
Specified the proposal as a table of airfoil, chord, angle, and leading-edge position for all 5 elements, referenced to the car's STL origin.

Downforce costs drag. This proposal stays inside the MY26 endplate and buys 30 N/m of downforce for 0.6 N/m of drag.

A 2D section isolates the element layout, which makes it a fast first screen before 3D runs with the endplates and the rest of the car.

Two 2D velocity contour plots of the five-element rear wing, MY26 and the MY27 proposal, with slow flow in blue and fast flow in orange and red — Velocity magnitude at 15 m/s around MY26 (left) and the MY27 proposal (right)

Overlay of the five MY27 rear wing elements, labeled E0 to E4, on the dotted MY26 elements inside the endplate outline — MY27 elements in color inside the MY26 endplate; dotted outlines are the MY26 elements

Bar chart of downforce per element for MY26 and the MY27 proposal; the main element and first lower flap gain the most — 2D downforce per element, MY26 against the MY27 proposal

MY27 element layout. s is distance downstream of the car's STL origin, u is height, and a positive angle means trailing edge up.
Element	Airfoil	Chord, mm	Angle, °	LE s, mm	LE u, mm
E0 main	S1223RTL	450	0	−293.0	852.1
E1 lower flap 1	S1223	228	33	146.0	874.6
E2 lower flap 2	S1223	142	58	321.5	1008.3
E3 upper front	S1223	200	6	−120.0	1030.1
E4 upper flap	S1223	160	30	71.3	1066.0

Active aero concept design review

MIT Motorsports, MY27 front wing, October 17, 2026

CoDR excerpt (PDF, 17 slides)

My role: Responsible engineer, active aero
Presented by: Me and 2 teammates
System: Servo-driven front-wing flaps, one servo per side
Chosen concept: Rod linkage
Next gates: Preliminary, then fabrication design review

Flap response: ≤ 0.2 starget 0.1
Positioning: ±5°target ±3°
Repeatability: ≤ 0.5°peak to peak
Durability: ≥ 1,000cycles
Actuator mass: ≤ 0.5 kgtarget 0.3
Power draw: ≤ 20 Wpeak

Functional requirements from the CoDR. These are design targets, not test results.

Set 23 functional requirements and broke down the 60+ FSAE 2027 rules that apply, so every flap position stays inside the front-wing envelope (no more than 700 mm ahead of the front tires or 500 mm high) and is designed to pass the inspectors' 200 N push test.
Compared 3 mounting locations and 2 linkage types with the team, scored two of the mounts side by side on 8 criteria from wing removal to flap slop and EE wiring, and chose a rod linkage for its lower mass, power draw, and complexity.
Budgeted the system at about $180 and 0.4 kg on the wing (1.42 kg with the wiring harness) with two 35 kg·cm servos, and designed it to fail safe: the flaps default to the high-downforce position, hard stops cap their travel, and a Feb 19 go/no-go keeps the wing passive if it isn't track-ready.

Active aero lets a wing change on track: more downforce in corners, less drag on straights. For MY27 the team chose the front wing, where moving a flap shifts the car's center of pressure, which helps new drivers. A servo on each side moves the flaps based on speed, steering, throttle, and brake.

MIT Motorsports hasn't run active aero on a car before. The review builds on earlier attempts: MY25's rear-wing flaps used an untested servo and a cable that bent too tightly at its pulley, and MY26's design lost more downforce than it saved in drag.

Next, the team will pick the best aero balance for each corner from a yaw moment diagram, then use CFD to find the flap angle that hits it.

CoDR slide explaining active aero, with a CFD image of the front wing — System overview: what active aero does, and its fail-safe

CoDR slide with front-view sketches of the MY26 wing and the concept's flap positions at low and high speed — The flaps at low and high speed, front view

CoDR slide with a hand sketch of the rod linkage driving the flaps from a servo — Chosen concept: rod linkage, section view

CoDR slide with a dimensioned hand drawing of the servo mount and the servo's torque table — Servo hand calcs and mounting

CoDR table of front-wing envelope requirements with targets and FSAE rule sources — Front-wing envelope requirements, traced to FSAE rules

CoDR pie chart of the active aero mass target by part, led by the servos — Target mass breakdown by part

Radiator heat-rejection model

MIT Motorsports, 2026

My role: Built and validated the model
Software: STAR-CCM+
Model: Dual-stream heat exchanger, porous media
Mesh shown: 6 regions, 49,632 cells
Feeds: Sidepods concept review

Validated the model against the radiator's air-side and coolant-side pressure drops (93 Pa and 5.1 kPa) and its outlet temperatures, by modeling the core as coupled air and coolant porous regions between inlet and outlet ducts and tanks.
Showed the stock core rejects 2.89 kW against the 2.34 kW worst-case requirement, a 24% margin, which justified cutting the core 8 mm thinner to save 0.5 kg in total.
Set the sidepods' air requirement at 0.17 kg/s per core (at a 3.4 m/s face velocity) from the same simulation.

MY27 runs two cooling loops, one radiator per sidepod. Each core has to reject 2.34 kW at worst-case endurance, 4.68 kW in total, a provisional figure from last year's data and inverter testing.

In the hand thermal-circuit check, with 0.18 kg/s of 30 °C air and 0.10 kg/s of 60 °C water, the air side limits heat transfer (181 W/K, about half the coolant side's capacity rate). That makes the sidepods' main job getting enough air through the core.

CAD model of the PWR radiator with its top tank and two coolant fittings — The PWR radiator, as CAD

Hexahedral mesh of the radiator model showing the air ducts, core, and coolant tanks — Six-region mesh: upstream and downstream air ducts, air and coolant cores, top and bottom coolant tanks

Radiator CAD shown in magenta overlaid on the radiator mesh — Radiator CAD (magenta) lined up inside the mesh

Temperature contour on the radiator core from 313 to 333 kelvin, hottest at the top — Core temperature, 313–333 K. Hottest at the top tank, where 60 °C (333 K) water enters.

Sidepods concept design review

MIT Motorsports, MY27 sidepods and radiators, October 2026

CoDR excerpt (PDF, 11 slides)

My role: Responsible engineer, sidepods
Presented by: 6 members of the aero subteam
Review: Concept design review (CoDR)
Decisions: Radiator mount, pod-shell mount, pod shape
Next gate: Preliminary design review

Set 7 functional requirements from the radiator model, the ArchDR, and the rules, including at least 2.34 kW and 0.17 kg/s of air per core, no more than 3 mm of deflection at 30 m/s, and zero added floor thickness, and listed the 9 FSAE 2027 rule constraints every concept has to meet.
Set a service target of 5 minutes or less per side, down from 15+ minutes on MY26, by splitting the radiator mount from the pod-shell mount so a sidepod comes off without opening the cooling loop.
Compared 3 mounting concepts (a frame-mounted duct, a duct on the floor and body panel, and stepped elements on the floor) on service time, floor attachment, frame tabs, and tooling, and queued 3 pod shapes for inlet CFD ahead of the preliminary design review.

The ArchDR treated the sidepods and radiator as one floor-mounted unit. The CoDR splits that into two decisions, the radiator mount and the pod-shell mount. Last year's shared mounting took more than 15 minutes to remove.

The thermal check says the air side limits heat rejection, which makes the deciding question for each pod shape whether enough air reaches the core. When the car is stopped, in the queue or during the driver change (up to 3 minutes under rule D.12.8.2), the fans take over.

CoDR slide on changes since the ArchDR and last year's design — What changed since the ArchDR and MY26

CoDR slide with requirements for heat rejection, air mass flow, aero package, and service time — Functional requirements, part 1

CoDR slide with requirements for mount stiffness, mass, and added floor thickness — Functional requirements, part 2

CoDR slide listing the rule constraints on the sidepods — Design constraints from the FSAE 2027 rules

CoDR critical path diagram from final heat load to detail CAD — Critical path to the preliminary design review

CoDR timeline of milestones from October through December — Timeline milestones

Rear-wing CFD training case

MIT Motorsports, aero team onboarding, 2026

My role: Built the case and ran the training
Trained: Full aero team, plus 7 new members
Geometry: 11-body rear-wing assembly
Mesh: 3.43 M cells, 10.27 M faces
Software: STAR-CCM+ 2606

Trained the full aero team on STAR-CCM+ and am onboarding 7 new members with this case, which takes a trainee from CAD import through meshing and solving to reading surface pressure.
Meshed the 11-body wing, endplate, and pylon assembly into a 3.43-million-cell air region with 10.27 million faces.

The aero team designs in STAR-CCM+, so new members need the workflow before they can contribute. This case gives them a real part to learn on, a multi-element rear wing with slotted endplates and pylons.

Rear wing CAD with three elements, slotted endplates, and two pylons, shown in STAR-CCM+ — The rear-wing assembly in STAR-CCM+'s 3D-CAD environment, 11 bodies

Surface pressure contour on the rear wing elements, endplate, and pylon — Surface pressure on the wing elements, endplate, and pylon

Turbine and wind tunnel

Two projects off the car: the turbine stage for MIT's student gas-turbine team, and wall corrections at the Wright Brothers Wind Tunnel.

Turbine stage: worksheet check, rotor CAD, blade FEA

MIT Sustainable Engine Team (SET), turbine team, October 2026

My role: Turbine team responsible engineer
Stage: Single-stage axial, 24 stator and 27 rotor blades
Job: Drive a 24 kW compressor at 159,000 rpm
Inlet: 0.1285 kg/s at 1033 K and 366.7 kPa
Tools: MATLAB, Onshape, ANSYS Mechanical

Caught 5 issues in the team's mean-line turbine worksheet, 3 calculation errors and 2 design risks, by writing a MATLAB check that rebuilds the stage from the sheet's inputs and matches its velocity triangles to within 0.01%.
Showed the stator hub exit Mach is 0.976, over the team's 0.95 limit, not the 0.952 the sheet reports, by computing Mach at each radius from that radius's own static temperature instead of the mean.
Found both blade rows outside the 0.75–0.85 Zweifel loading band (stator 0.90, rotor 0.58) and the 24/27 blade counts sharing a factor of 3, a forced-response risk, then swept blade count to find 4 pairs, such as 27/20, that fix both to first order.
Modeled the bladed rotor in Onshape with an M8 × 1.0 threaded hub, and ran a first-pass ANSYS FEA of one Haynes 230 blade under 100–500 kPa gas loads from 27 to 702 °C: peak von Mises stress of 30.9 MPa, about 9× below the roughly 270 MPa yield at 650 °C under gas loads alone.

SET is MIT's student gas-turbine team. Its turbine is a single stage that takes 0.1285 kg/s of 1033 K gas and extracts the 24 kW that spins the compressor at 159,000 rpm. The team's mean-line worksheet sizes the stage's velocity triangles, flow path, and blade losses for a predicted 80% efficiency.

A mean-line sheet can break quietly: one cell uses the wrong temperature or radius, and every number downstream still looks plausible. The check is built to catch that, and it can be rerun as the design changes.

Worksheet check in MATLAB

The script reads each input from a named worksheet cell, recomputes the mean line, the hub-to-tip velocity triangles, the stator exit state, and the blade loading, then compares each result with the worksheet's own output.

What the check found. Blade-count options are first-order, at fixed axial chord; the losses have to be rerun in the worksheet for any new count.
Finding	Worksheet	MATLAB check
Stator hub exit Mach, limit 0.95	0.952, from the mean static temperature	0.976, from the hub's own 888 K
Stator exit static temperature	935 K, mixing two c_p values	921–923 K with one c_p
Rotor tip inlet relative angle	−5.3°, mixing two radii	−0.9° at one radius
Zweifel loading, target 0.75–0.85	Stator 0.90, rotor 0.57	In band at 26–28 stator and 19–20 rotor blades
Blade counts	24 and 27 share a factor of 3 (forced-response risk)	In band with no common factor: 26/19, 27/19, 27/20, 28/19

Two MATLAB plots against span: stator exit Mach number, where the local-temperature curve starts at 0.976 above the 0.95 limit, and static temperature, which rises from 888 K at the hub — Stator exit Mach and static temperature from hub to tip. With each radius's own static temperature, the hub reaches Mach 0.976.

MATLAB plot of Zweifel coefficient against number of blades for the stator and rotor, with the current design points marked outside the target band — Zweifel loading against blade count at fixed axial chord. The current 24 stator and 27 rotor blades sit outside the 0.75–0.85 band.

Six MATLAB velocity-triangle plots for the rotor inlet and exit at hub, mean, and tip, showing absolute, relative, and blade velocities — Rotor velocity triangles at hub, mean, and tip. At the tip inlet, the worksheet's dashed vector comes from mixing two radii.

Stacked bar chart of loss coefficients: stator 0.069 total, rotor 0.346 total, mostly tip leakage — Loss breakdown from the worksheet's loss model: tip leakage is 81% of the rotor's loss coefficient of 0.346.

Rotor CAD and blade FEA

The rotor is a bladed disk in Onshape with an M8 × 1.0 threaded hub, 15 mm deep, for the shaft; airfoil blades from the worksheet's metal angles come next. For the FEA, one blade is a 10 × 5 × 2 mm Haynes 230 bar fixed at the root, with 500 kPa on one face and 100 kPa on the other four, swept from 27 to 702 °C in 15 steps. This first pass carries gas loads only; the centrifugal load at 159,000 rpm is the next step.

Onshape model of the turbine rotor, a disk ringed with rectangular blades and a central threaded bore, beside its feature tree — Rotor in Onshape: a bladed disk with an M8 × 1.0 threaded hub, 15 mm deep. The blades are still blocks.

ANSYS contour of von Mises stress on the blade bar, highest near the fixed root — Von Mises stress: 30.9 MPa peak, at the root

ANSYS contour of total deformation on the blade bar, zero at the fixed root and largest at the free tip — Total deformation: 4.3 µm at the tip

ANSYS contour of normal stress along the blade, tension on one face and compression on the other near the root — Bending stress along the blade: ±38 MPa at the root

Wall corrections for the Wright Brothers Wind Tunnel

UROP with Prof. Mark Drela, MIT AeroAstro, since September 2026

My role: UROP research assistant
Test section: Closed, 12 × 7.75 × 18 ft, 15–230 mph
Code: MATLAB and XFOIL, on GitHub, documented in MkDocs
Reference: Barlow, Rae & Pope, chapters 9–11
Cross-check: Drela, Flight Vehicle Aerodynamics, ch. 10

Implementing wall-interference and blockage corrections in MATLAB, with XFOIL, to make the force and pressure data the tunnel delivers to industry, research, and student tests more accurate, and documenting the correction code in MkDocs.
Marked up the 108-page boundary-corrections section of Barlow, Rae & Pope's Low-Speed Wind Tunnel Testing for the tunnel's code: 275 highlights on 82 pages, sorted into procedure (55), equations that apply to the WBWT (137), and validity checks (83), cross-checked against Drela's Flight Vehicle Aerodynamics.
Collected ready-made unit tests for the correction code from the book's worked examples, including an 8 × 12 ft tunnel with corner fillets, the closest match to the WBWT in the book.
Leading my own airfoil project from CAD through CFD to a tunnel test, to compare predicted and measured lift, drag, and pressure across 10–15 angles of attack.

Walls change the flow around a model. In a closed test section like the WBWT's, the model and its wake squeeze the flow, so it speeds up around the model, and the walls cut the downwash, so lift reads high and drag reads low. Classical corrections estimate each effect separately and add them up.

The open question for the WBWT is its corner fillets. The book's image methods model a pure rectangle, so the code will need either a panel model of the real section or a measured check.

WBWT test section looking downstream, to scale, drawn from the tunnel's published specifications. The corner fillets taper along the 18 ft section, opening the flow area slightly to offset wall boundary-layer growth.

What the corrections fix in a closed test section
Effect	What it does	Corrected through
Solid blockage	The model's volume speeds up the flow around it	ε_sb, raises dynamic pressure
Wake blockage	The wake does the same and adds drag	ε_wb, raises dynamic pressure
Streamline curvature	The model acts more cambered, so lift, moment, and angle read high	Δα, ΔC_L, ΔC_m
Lift interference	The walls cut the downwash, so lift reads high and drag low	Δα = δ(S/C)C_L ΔC_D = δ(S/C)C_L²
Buoyancy	A pressure gradient along the section pushes on the model	The tunnel's centerline static survey

Build and test

I spent my first year on MIT Motorsports on the composites team. I fabricated carbon-fiber parts for 4 vehicle components (body panels, wing elements, the seat shell, and the battery enclosure), each about 15% lighter than the previous year's, and machined 7 molds, tooling fixtures, and parts on manual mills, lathes, and the waterjet.

Rear wing mounting retrofitSolo

Aero mounting

I retrofitted the rear wing's mounting on my own. I worked out how the wing attaches to the chassis, then designed and fabricated the spacers that set its installed height and angle.

Rear wing mounted on the MIT Motorsports car in the shop — Rear wing installed on the car

Carbon strut connecting the rear wing mount to the chassis — Carbon strut tying the wing mount into the chassis

Wing elements, foam plug to prepreg

Composites, aero surfaces

Each element starts as a foam plug shaped to the airfoil and sealed with epoxy, so resin can't soak in and the cured part releases cleanly. Any flaw on the plug ends up on the wing. The elements are then laid up in prepreg, bagged, and cured.

Pink foam plugs for wing elements, sealed and ready for layup — Sealed foam plugs for the wing elements

Blue-backed prepreg laid over a wing mold — Prepreg laid into the mold before bagging

Cured carbon fiber wing element on a workbench — Cured carbon wing element after demolding

Large composite body panel

Composites, body panels

A large panel laid up over sealed foam molds and vacuum-bagged in two sections. After the cure, the foam bonded to the inside has to be dug out and the flanges trimmed before the part fits the car.

Foam molds under vacuum bag during cure — Molds under vacuum during the cure

Cured carbon panel with pink foam still attached — Cured panel with foam still bonded to the inside

Trimmed carbon body panel — Panel after foam removal and edge trimming

Layup characterization

Materials testing, MIT D-Lab

The team made 6 test panels that vary fabric, ply count (3 to 5), orientation, and cure schedule, logging the carbon-to-resin ratio before and after cure; final ratios ran from roughly 50:50 to 83:17. I tested coupons in tension and compression on an Instron to check strength and stiffness against the layup requirements.

Carbon fiber test panels with handwritten labels listing layup and cure details — Test panels, each labeled with fabric, plies, orientation, cure, and carbon-to-resin ratio

Cured flat carbon panel — Cured panel before cutting coupons

Instron screen showing a force versus displacement curve — Force–displacement trace from a coupon test

Impact attenuator crush testing

Structures testing

I supported the team's crush tests. FSAE requires the attenuator to absorb a set amount of energy without exceeding a deceleration limit, shown by test. A thin-wall tube and an aluminum honeycomb block bonded to the anti-intrusion plate were crushed on an Instron, and the area under each force–displacement curve gives the energy absorbed.

Aluminum honeycomb specimen in an Instron test frame — Honeycomb specimen set up for a compression run

Metal tube being crushed in an Instron — Thin-wall tube under lateral crush

Crushed honeycomb bonded to an aluminum plate after testing — Crushed honeycomb bonded to the anti-intrusion plate

Carbon-fiber seat shell

Composites

A one-off seat laid up over a foam plug sanded to the driver's fit and sealed with epoxy. It has to hold the driver where the cockpit was designed around and survive a full event.

Foam seat plug sealed with epoxy — Foam plug sealed and sanded before layup

Cured carbon seat shell with foam being removed — Cured shell after demolding, residual foam being cleared

Stacked dry carbon fabric on a table — Dry carbon fabric staged for cutting

Battery enclosure

Composites, EV packaging

The accumulator container: a bonded carbon box with aluminum edge reinforcement, split into segments with copper foil routed between them. It has to carry the pack, meet the enclosure and isolation rules, and fit the space the chassis leaves.

Open composite battery box with internal walls lined with copper foil — Segment walls and copper foil routing before closeout

Closed carbon battery enclosure with aluminum edges — Bonded enclosure with aluminum edge reinforcement

Earlier work

Custom LEGO accessories in SolidWorks

High school senior project, fall 2024

My role: CAD assembly and drawings
Team: 3 students
Software: SolidWorks
Base set: LEGO Creator 3-in-1 Space Shuttle (31134)
Deliverables: 15-slide product pitch, dimensioned drawing

Modeled the set's spaceship build as a SolidWorks assembly, then designed 3 new accessories with my team that attach to it: engine flames, wing-mounted laser cannons, and a flag that rotates 360°.
Drew the flag sub-assembly at 3:1 scale in ABS, dimensioned to LEGO's 7.8 mm 1×1 footprint with a 30 mm pole.

A product-design exercise: take an existing set, model it, and propose accessories that fit both LEGO's system and the set's story. Each laser cannon is a single part, a 2×2 base with six extruded features that rise 84.3 mm above the connector.

The flag sub-assembly drawing is mine, dated November 25, 2024.

SolidWorks render of the LEGO spaceship with two wing-mounted laser cannons — Spaceship assembly with the laser cannons

SolidWorks render of the LEGO spaceship with orange engine flames — With the engine flames

SolidWorks render of the LEGO spaceship with a flag mounted on it — With the flag

Dimensioned engineering drawing of the flag sub-assembly with title block — Flag sub-assembly drawing, 3:1 scale

SolidWorks render of the laser cannon part — Laser cannon, a single part on a 2×2 base

SolidWorks render of the orange flame part — Flame accessory part

Experience

MIT Motorsports

Formula SAE Electric, Edgerton Center (N52)

Jun 2026 – present

Aerodynamics responsible engineer: rear wing, active aero, sidepods

Built and validated a STAR-CCM+ radiator model against 93 Pa air-side and 5.1 kPa coolant-side pressure drops; it showed a 24% heat-rejection margin over the 2.34 kW requirement and justified an 8 mm thinner core that saves 0.5 kg in total.
Direct the MY27 aero package team's design meetings, scoping a full redesign toward C_LA 4.0 ± 0.2 m², C_DA 1.3 ± 0.1 m², and a 42/58 balance within a 14.22 kg, $4,100 budget under the FSAE 2027 rules.
Trained the full aero team on STAR-CCM+ workflows, and lead onboarding and training for 7 new members.
Updating the rear wing (a 2D layout study shows 9% more downforce for 3% more drag), designing new sidepods, and bringing active aero to the front wing, which the team hasn't run on a car before.

Aug 2025 – May 2026

Composites team responsible engineer

Fabricated carbon-fiber parts for 4 vehicle components (body panels, wing elements, seat shell, battery enclosure), each about 15% lighter than the previous year's, using prepreg and wet layup, vacuum bagging, and bonding.
Tested carbon-fiber coupons in tension and compression on an Instron at MIT D-Lab to validate layup schedules against design requirements, and supported impact attenuator crush testing.
Machined 7 molds, tooling fixtures, and parts on manual mills, lathes, and the waterjet, and prepared foam plugs for molds.

MIT Sustainable Engine Team (SET)

Student gas-turbine team, Energy Commons

Oct 2026 – present

Turbine team responsible engineer

Modeled a bladed-disk turbine rotor with an M8 × 1.0 threaded hub for the shaft. The stage drives a 24 kW compressor at 159,000 rpm, and the team's mean-line model sizes its velocity triangles, flow path, and blade losses for 80% efficiency.
Built a MATLAB check of the mean-line worksheet that reproduces its velocity triangles and caught 5 issues, including an understated stator hub exit Mach (0.952 in the sheet, 0.976 against a 0.95 limit) and a 24/27 blade count with forced-response risk.
Ran a first-pass ANSYS FEA of a Haynes 230 blade under 100–500 kPa gas loads from 27 to 702 °C, finding peak root bending of ±38 MPa and about a 9× margin to yield at 650 °C under gas loads alone.

Wright Brothers Wind Tunnel, MIT AeroAstro

With Prof. Mark Drela, creator of XFOIL

Sep 2026 – present

UROP research assistant

Implementing wall-interference and blockage corrections in MATLAB, with XFOIL and GitHub, to improve the accuracy of force and pressure data for industry, research, and student test campaigns.
Lead student and industry tunnel projects, from developing and installing their models to post-processing the data.
Leading an independent airfoil project from CAD through CFD to a tunnel test, comparing predicted and measured lift, drag, and pressure distribution across 10–15 angles of attack.
Marked up the boundary-correction chapters of Barlow, Rae & Pope (275 highlights on 82 pages) as the reference for the tunnel's correction code.

realtalk@MIT, Center for Constructive Communication

Cambridge, MA

Jun 2026 – Sep 2026

UROP research assistant

Coded 11 conversation transcripts with a structured qualitative codebook and identified 33 cross-transcript themes.
Wrote 6 analytical memos and thematic summaries that turned raw transcripts into structured findings for the research lead.
Built participant-demographic visualization slides for a conference presentation, and ran an outreach campaign to 15 participants.

Leadership and campus

MIT AIAA Student Chapter

Sep 2026 – present

Executive board member

Organizing technical talks, industry panels, and events for the aerospace student community and faculty.

MIT First-Year Orientation

Apr 2026 – present

Orientation leader and associate advisor

Guiding incoming first-years through orientation and their transition to campus life.

MacGregor House Makerspace

Feb 2026 – Aug 2026

Makerspace mentor

Trained residents on 3D printer setup while maintaining hardware, managing inventory, and helping optimize models.

Skills

Simulation and analysis

STAR-CCM+ external aero, porous-media heat exchangers
ANSYS Fluent
ANSYS Mechanical static structural FEA
XFOIL
MATLAB, Python mean-line turbine checks
HPC on MIT's Engaging cluster

CAD, drawings, and code

SolidWorks parts, assemblies, drawings
Onshape
Siemens NX and Teamcenter
GD&T
GitHub, VS Code, MkDocs

Composites

Prepreg and wet layup
Vacuum bagging
Foam plugs and molds
Bonding and finishing

Machining and fabrication

Manual mills and lathes
Waterjet
Welding
3D printing
Drill press, band saw, power tools

Testing and reviews

Tension and compression testing (Instron)
Crush testing
Wind tunnel testing
Design reviews: ArchDR, CoDR
FSAE 2027 aero rules

Certifications and languages

AE1110x Introduction to Aeronautical Engineering TU Delft
Microsoft Office Specialist: Excel Associate 2023
Microsoft Office Specialist: Word Associate 2022
English and Spanish native

Education

Massachusetts Institute of Technology

Cambridge, MA

Aug 2025 – May 2029

B.S. Aerospace Engineering (Course 16), GPA 4.75 / 5.0

Coursework: Materials and Structures, Thermodynamics and Propulsion, Fluid Dynamics, Signals and Systems, Principles of Automatic Control, Differential Equations, Intro to Computer Science and Programming in Python, Statistics and Probability, Physics I and II.

South Texas College

McAllen, TX

Aug 2022 – Jun 2025

Associate of Interdisciplinary Science, Mechanical Engineering major, GPA 3.87 / 4.0

Contact

I'm looking for summer 2027 internships in aerodynamics, CFD, FEA, and propulsion, in aircraft, engines, or motorsports. Email is the fastest way to reach me.

eddy215@mit.edu

Scan to save my contact card