Skip to content
Back to home
PRODUCT · Taishu·Design

From “slab design” to
“floor structural-system design”

What's truly hard about industrial floors isn't slab thickness — it's subgrade, interface friction, voids and long-term prestress: the parts most reliant on experience and hardest to standardize. Taishu·Design makes them reviewable, traceable and engineer-signed.

4
engineering cores
100%
traceable conclusions
Range
low-confidence params
Taishu·Design · Engineer Workbench
DEMO
Slab thickness260 mm
Effective prestress (end/mid/far)1.30 / 1.15 / 0.95 MPa
Subgrade sensitivityHigh
Review triggerNon-uniform subgrade · contact/void
Formal output is reviewed, signed and archived by the engineer

The system doesn't replace engineers with a black box; it turns the parts of floor design that most rely on experience, are most easily underestimated and hardest to standardize, into reviewable, traceable, continuously improving technical assets.

CORE TECHNICAL BARRIERS

Why plain slab-thickness calculation
can't manage an industrial floor

A floor's real risk depends on the interaction between slab, support system, interface system and long-term operation. Taishu·Design upgrades the “floor slab” to a “floor structural system”.

01

Subgrade effects

The subgrade isn't a single k value. Bearing capacity, settlement, differential settlement, uneven fill, trenches, groundwater, cold-storage frost heave, long-term degradation and slab voids must all be handled.

Core question: can the subgrade provide stable, uniform, continuous, predictable support throughout the design life.

02

Interface friction & slip

The horizontal direction governs prestress release and shrinkage-crack risk; the vertical direction governs bending-tensile amplification under wheel loads, rack posts and equipment bases.

Goal: controlled low friction, reliable vertical support, necessary edge release, key-point restraint.

03

Long-term · local heavy load · nodes · construction

Shrinkage/temperature/creep/relaxation/anchorage and friction losses together set long-term effective prestress; forklifts/AGV/rack posts bring fatigue; column edges, doorways, temperature-zone joints and construction-stage control decide real crack risk.

These “most underestimated” parts are exactly what decide whether a floor serves long-term.

FOUR-LAYER CORE

How fast routine projects ship,
how complex ones get reviewed

Four cores with clear roles: layers 1, 2 and 4 form the main delivery loop for routine projects; the third high-order FE layer is a conditionally-triggered, templated review.

01Decision hub

Engineering Rule Engine

Codifies standards, company experience, scenario differences and risk judgment into executable rules. Even with sparse input it forms a sound path and clearly surfaces data gaps and items needing review.

02Main design for routine projects

Light Calculation Core

Explainable calculation of slab-subgrade, prestress, loads, fatigue and subgrade sensitivity. More reliable than Excel, lighter than general FE; outputs split into deterministic, range, risk grade and review trigger.

03Complex cases · conditional

High-order FE Review Core

Templated review for non-uniform subgrade, contact/void, friction slip, local overloads and cold-storage repair. Entered only when review triggers are hit — never overused.

04Customer-perceived value

CAD/BIM Auto Drawing & Delivery

Turns results into construction base drawings, node drafts, calc reports, material lists and asset-archive data (DXF/PDF/Word/IFC) — decoupled from the calc core, version-traceable.

Main delivery loop = rule engine + light calc + auto drawing; high-order FE is templated review evidence for complex cases.
ENGINEERING CREDIBILITY

Who's accountable when issues arise?
Every conclusion is traceable to its source

Every design conclusion carries five elements — low-certainty conclusions use range, grade or review trigger, and engineer confirmation is completed before formal delivery.

1

Parameter source

Client input / drawing recognition / default assumption / engineer confirmation / rule derivation, each labeled.

2

Confidence level

Low-confidence params (settlement/friction/void) use range and grade — no fake precision.

3

Calculation method

Each check shows formula, values, result, conclusion and governing case.

4

Risk explanation

Each risk item gives a recommendation and whether it enters high-order review.

5

Engineer sign-off

Formal output is reviewed, signed and archived by the engineer; AI only extracts and drafts, never decides structural parameters.

The evidence chain uses append-only records + hash integrity checks: conclusions can't be altered after the fact and can be verified unchanged.
DELIVERABLES

What customers buy
is executable engineering output

Engineering Scheme

Floor-system recommendation, initial parameters, risk grade and option-comparison basis.

Structural Calc Report

Generated from structured results as the single source; sources, formulas and conclusions traceable, versions bound to rules/core.

Drawings · Nodes

Parametric base drawings and templated node drafts, completed by engineers into construction details, entering internal review and release.

Asset Archive Data

Long-term service risk and asset-archive data (incl. IFC), can flow into Taishu·Twin for a design-operation loop.

USE CASES

For the long-term value of industrial buildings

Cold StoragePrecision Clean ManufacturingAutomated High-Bay WarehousingContinuous ProductionIndoor Heavy-Load TrafficOutdoor Heavy-Load Yards

The design baseline flows into the operation archive

Parameters and drawings frozen by Design become the baseline for acceptance and inspection in Taishu·Twin — design and operation, one set of engineering objects, one evidence chain.

Explore Taishu·Twin

Hand the hardest floor structural design
to a reviewable engineering core

Whether new prestressed floors, heavy-load plants or cold-storage repair, start with one technical talk.

Online Chat