Vorsprung Autowerk

Uberbrick Rev F

A water-to-air charge cooler for the Audi 3.0T supercharged V6 (EA837 with the TVS1320 supercharger).

The design is final. We are vetting manufacturers now and plan to order prototypes in December.

How it compares with the OEM brick on an 80 °F day

We ran the Rev F design and a factory-style brick through the same thermal model, with the same airflow and the same coolant. Drag the slider to change how hard the engine is pulling.

Full pull at 24 psi. Airflow per brick: 0.23 kg/s.

OEM brick, air leaving
152 °F (67 °C)
Uberbrick, air leaving
139 °F (60 °C)
The Uberbrick leaves the air 13 °F cooler.
Room before the ECU cuts power
The cut happens at 176 °F (80 °C).
OEM brick: 24 °F. Uberbrick: 37 °F.
Air leaving the brick, by boost
100
120
140
160
180 °F
4
8
12
16
20
24 psi
ECU power cut, 176 °F (80 °C)
OEM brick
Uberbrick

At cruise, both bricks leave the air about 25 °F above the ambient temperature. The gap opens as the engine works harder, and at a full 24 psi pull the model puts the Uberbrick 13 °F cooler.

How these numbers were produced
  • Ambient is 80 °F (27 °C). Coolant enters the bricks at about 103 °F on a pull, a figure consistent with cars that have an upgraded heat exchanger.
  • Supercharger efficiency is 60%. Coolant flow is 45 L/min across both bricks. Airflow per brick rises from 0.06 kg/s at cruise to 0.23 kg/s at 24 psi.
  • The factory brick is scaled so it reads about 149 °F (65 °C) on a hard pull at 77 °F, the middle of the 140 to 158 °F that these cars typically show. The same scale is applied to the Uberbrick.
  • Treat any single number as an estimate, roughly 15 °F either way. The gap between the two bricks is more reliable than either number.
  • This is a simulation, not a measurement.

The engineering behind the numbers

Two schematics showing how a brick like this actually moves heat. These are illustrations of the general construction, not the Rev F build spec — the exact layer count, fin pitch and plate thickness stay confidential until the brick is in production.

Anatomy of a bar-and-plate core

Air and coolant run through alternating layers, brazed together with thin plates between them. Air passes through horizontally; coolant runs the other way through its own layers.

Cap plate Air channel Coolant channel Separator Pattern repeats through the stack · exact layer count is part of the build spec (schematic — not to scale)

Why fin geometry matters

A plain, continuous fin lets a layer of still air build up along its surface, and that layer insulates the metal from the airstream. An offset strip fin is cut into short, staggered segments, so a fresh leading edge appears every few millimeters and that insulating layer keeps getting knocked down and restarted. More of the metal stays in contact with moving air, and more heat gets out.

Plain, continuous fin Boundary layer keeps growing Offset strip fin Each edge restarts it

What happened, and what we changed

Our original design, which we call Rev.1, worked and performed like the OEM brick. For production, the manufacturer changed how the core was built. We later learned the job had been passed to another factory at a lower cost.

What came back, which we call Batch Rev.1, used a different fin than we specified. It ran much hotter, with poor intake air temperatures, and the original construction was no longer available. We stopped rather than ship parts we do not trust.

So we redesigned the brick from the ground up. The goal is a brick that beats the original, not one that only matches it.

What is different this time

  • Every build detail is confirmed in writing before any tooling is paid for.
  • We verify who actually owns the furnace and builds the brick.
  • The first bricks are sectioned and tested at an independent lab.
  • The prototype is tested against an OEM brick before production starts.

The redesign, in numbers

Rev.1 and Batch Rev.1 are not simulations built on paper. Rev.1 is the original design we built and tested. Batch Rev.1 is the production unit we tested after the factory substituted a different fin than we specified. We measured both directly and ran that geometry through the same thermal model used throughout this page. The bars on the right are manufacturers we are vetting for the redesign today.

In testing, Batch Rev.1's intake air reached the ECU's 176 °F (80 °C) protective cutoff, shown here as the bar that crosses the red line. That is a measured outcome, not a modeled one: direct evidence that its fin could not move enough heat into the coolant.

Previous factory-style assumption, 152 °F
ECU power cut, 176 °F (80 °C)
137.6 °F
58.7 °C
139.4 °F
59.6 °C
142.8 °F
61.5 °C
148.2 °F
64.5 °C
176 °F+
80 °C · hit and crossed the ECU cutoff
161.2 °F
71.8 °C
Supplier 3
Redesign quote,
tooled to spec (proposed)
Supplier 1
Redesign quote,
builds to spec exactly
Supplier 2
Redesign quote,
as confirmed today
Rev.1
Our original design,
built and tested
Batch Rev.1
What the factory built,
tested after production
Supplier 3
Same supplier's
standard part (not selected)
Confirmed in writing by the manufacturer
Proposed, not yet confirmed
Measured from the unit we built and tested
Confirmed by real-world testing, hit the ECU cutoff
Confirmed, but not the option we chose
How this comparison was built
  • Rev.1: fin geometry measured directly from the original design we built and tested, run through the same thermal model and the same 24 psi, 80 °F hard-pull condition as every other number on this page.
  • Batch Rev.1: fin geometry alone predicts about 164 °F. In testing, the unit ran hotter and reached the ECU's 176 °F (80 °C) cutoff. We show the tested result here, because it is the more direct evidence of how little heat that fin was moving into the coolant. The bar is drawn crossing the cutoff line to make that unambiguous; we did not measure how far past 176 °F it actually went, since the ECU intervenes at that point.
  • The redesign quotes: built from the exact fin, plate and tank geometry each manufacturer has put in writing. Where a manufacturer has proposed something but not yet confirmed it in writing, the bar is marked proposed.
  • Manufacturer names are withheld while we finalize terms. We refer to them as Supplier 1, 2 and 3.
  • Rev.1 and Batch Rev.1 are two separate units, built and tested separately, not the same brick measured twice.

Where we are and what comes next

Dates are targets, not promises. If a test fails, the date moves and we will say so straight away.

  1. Redesign
    Done
    The design is final.
  2. Manufacturer vetting
    In progress
    Checking each manufacturer in writing.
  3. Prototype order
    Target: December
  4. Prototypes arrive
    Target: January
  5. Testing against an OEM brick
    January
  6. Production and delivery
    Target: late January to February

Next update by December 15, and again when the prototypes ship.

Fit and layout

Replaces OEM 06E145621S
Per car: Two bricks, one per cylinder bank
Ports: Two 3/4 in hose barbs, both on the front tank
Return end: Flanged tank with an O-ring face seal
Core size: 68 × 58 × 250 mm
Overall length: 355 mm with the barbs

The internal construction of the core is confidential until the brick is in production.

Vorsprung Autowerk

Drive the difference. Drive Vorsprung.

The temperature comparison is a simulation scaled to typical factory-brick behavior. It is not a measurement. Prototype testing against an OEM brick will show how close it is. Renders are illustrations, not production parts.