Requirements review
Ask MixLab about this recipe
Built in assistant. Answers from this model and reruns it for what if questions.A starting point for a trial batch, not a stamped design. Confirm with cylinder breaks.
Why trust this recipe
Model accuracy
Closest tested mix
Safety margin
Strength over time
Slag and fly ash react slowly, so the greener mix starts weaker and catches up by the test age.
How it works
- Predict. Five gradient boosted tree models, trained in your browser on 1,030 measured lab tests, predict strength from the ingredients and age.
- Stay inside real data. Every candidate is exactly 1 m³, uses amounts seen in real tests, and must sit close to a recipe that was actually tested.
- Add the margin. The mix must reach f′c + 1.34σ, the ACI 318 approach, where σ combines tested model error and model disagreement.
- Search. About 9,000 recipes are tried. MixLab returns the lowest carbon one that clears the margin, next to the leanest cement only mix.
CO₂ and price assumptions
Typical CO₂ factors on the order of the ICE database. Prices are illustrative, so enter local ones. The recipe updates when you change a value.
Limits
The data comes from one lab's family of materials, and local materials differ. Workability and durability are not predicted. Always confirm with a trial batch.
Sources
Data: I-Cheng Yeh, Cement and Concrete Research 28(12), 1998, UCI Concrete Compressive Strength dataset. Required average strength: ACI 318. Carbon factors: order of magnitude of the ICE database (University of Bath). Cement is about 8% of global CO₂: Chatham House, Making Concrete Change, 2018.