Welcome to Guangdong Qinglong Construction Engineering Co., Ltd.!

Site Map   |  Contact Qinglong   | 

English
  • 中文
  • English
  • عربي
Qinglong UHPC & GRC

UHPC & GRC Complex Architecture Manufacturing

A Global Benchmark in Smart Architectural Fabrication

400-777-2203
0760-88706348
13902597531
Image display
资讯中心

2026 UHPC Ultra-High Performance Concrete Engineering Quality: Why a Project Can Still Fail Even When Test Cubes Pass

2026-11-18 11:53:07

Click:

A common scenario appears in project acceptance documents: the 28-day compressive strength report for a UHPC test cube is qualified, and the component appearance is acceptable, yet half a year after delivery, network cracks appear in the bridge deck paving, or the joints of curtain wall hanging panels begin to seep water. All parties dig out their documents and cross-check—the material report is fine, and the construction records are fine, so where does the problem lie.

This scenario reveals a fact repeatedly validated in UHPC projects: passing material inspection does not equal passing engineering quality. UHPC engineering quality is jointly determined by four layers—material grade, structural design, process execution, and process documentation. A test cube report covers only the material layer; a problem in any of the other three layers will manifest after delivery in the form of cracks, leakage, chipped edges, rust stains, and so on. Following this framework, this article breaks down what actually constitutes UHPC engineering quality in 2026, who is responsible, and what documentation proves it.

1. First, Distinguish Three Easily Confused Meanings of 'Quality'

To discuss UHPC engineering quality, three terms with different meanings must first be distinguished:

• Material quality: whether the performance of the fresh mix and hardened body reaches the agreed grade, judged by standards and test reports;

• Structural quality: whether design configurations such as connections, joints, and anchoring match the load-bearing logic, judged by design documents and calculation sheets;

• Engineering quality: the comprehensive result of the above two layers plus process execution and process records, judged by the acceptance system.

Online articles often confuse the three—using material compressive strength data to argue 'high engineering quality,' which does not hold logically. A UHPC bridge deck paving may crack even though the material grade itself meets the standard, with the problem lying in the curing regime or the paving process; conversely, material batch fluctuations can indeed render a well-executed construction scheme meaningless. The reason responsibility division matters is that defects at different layers belong to different responsible parties.

2. Overview of the Four-Layer Definition of Engineering Quality

Layer

Determining Factors

Main Responsible Party

Basis for Judgment

Typical Failure Form

Material layer

Mix proportion, raw materials, grade determination

Production unit

Product standard + test report

Insufficient strength, inadequate durability

Structural layer

Connection configuration, joint design, load path

Design unit (with detailed design unit assisting)

Design documents, calculation sheets

Joint leakage, anchoring failure, excessive deformation

Process layer

Mixing, forming, curing, paving and finishing

Production / construction unit

Special scheme + process records

Fiber balling, plastic shrinkage cracks, chipped edges

Documentation layer

Testing commissioning, witness sampling, acceptance records

Construction entity leading all parties

Compliant documentation chain

Broken traceability, responsibility cannot be determined

 

The four layers are in an 'AND' relationship—if any layer is missing, overall quality cannot be established. The following expands layer by layer.

3. Material Layer: Grade Determination and Test Prerequisites

For the quality determination of the material layer, three things deserve special attention in 2026.

First, the grade criterion must be written together with the standard version. UHPC strength grades currently exist in a state where three grading systems coexist: the current national standard GB/T 31387—2015 with RPC100~180 (lower-limit values), the association standard T/CCPA 7—2018 with UC120/150/180 (interval values), and the revised draft-for-approval with UC100~200 (standard values, 95% guarantee rate). The revised version remained under review on the official platform as of November 2026, and some materials have already cited it under the number 'GB/T 31387—2025'—before an official release announcement is available, this is a citation of a pre-release number that cannot serve as a basis for acceptance. If a contract writes a grade without the standard version number, both sides can justify themselves in a dispute.

Second, the test age criterion must be fixed in the contract. The common criterion for test ages of UHPC exterior wall panel type products is natural curing not less than 28 days and steam curing not less than 7 days. Passing at 7 days of steam curing is not the same thing as passing at 28 days of natural curing; confusing the criteria is a common source of strength disputes.

Third, a grade is an indicator set, not a single number. Compressive strength is only one axis of grade determination; tensile (flexural) strength, elastic modulus, and durability thresholds jointly constitute the determination. 'Without fiber it is not UHPC,' but adding enough fiber does not equal reaching the grade—only by judging the whole set of indicators item by item, with the report attached with full data, is the quality proof of the material layer complete.

4. Structural Layer: Quality Is Locked In at the Design Stage

A common feature of structural layer quality defects is: they cannot be seen during construction, and appear concentrated after delivery.

Taking exterior wall hanging panels as an example, the industry's common load-bearing logic is to separate the bearing points (carrying the panel's self-weight) from the wind-resistance connection points (carrying wind load), provide elastic shims between hangers and the panel back to avoid rigid contact, and divide panel joints into two routes—closed joints and open joints: closed joints rely on backing rods to control the width-to-depth ratio and avoid three-sided adhesion of the sealant, while open joints rely on an equal-pressure drainage configuration. If these configurations are not locked in at the design stage, no leveling action on site can save them.

Structural layer quality proof materials include: connection node calculation sheets, joint schemes, and embedded-part positioning drawings. In the technical clauses of the contract, the purchaser should require the design or detailed design unit to submit these documents, and list 'configuration consistent with the calculation sheet' as a check item for entry acceptance—this is also an extension of the requirement in the 2026 policy context that 'the design unit shall state material technical parameters in the design documents': if the design documents give no parameters, the grade of the material layer has nothing to be compared against.

5. Process Layer: Two High-Risk Variables in the Factory Links

The roots of process layer quality defects are mostly in the factory; the site merely bears the consequences. Two high-risk variables deserve separate explanation.

Mixing and feeding. UHPC mix is sensitive to feeding sequence: the dry-mixing stage must use forced mixing equipment to achieve uniform dispersion of sand and cementitious materials, no water may be added midway in the wet-mixing stage, and steel fibers must be sprinkled in slowly at the final stage while maintaining low-speed mixing. Omitting any link causes at least a loss of flowability and at worst fiber balling—balling can be observed at the mix stage, but once poured and formed it can only be checked by sectioning, and the disposal cost differs by an order of magnitude.

Curing regime. The common steam-curing curve consists of four stages: standing, heating, constant temperature, and cooling; the heating and cooling rates and the temperature difference between the interior and exterior of the component are two hard constraints, and before leaving the shed it should be slowly cooled to an acceptable temperature difference with the environment—sudden cooling of thin-walled components after leaving the shed is a high-frequency inducement of cracking. Curing records (temperature and humidity curves, data from temperature-measuring elements) should be preserved as part of the factory documentation chain; when strength or crack disputes arise, the curing records are the basis for defining which link is responsible.

6. Site Layer: Plastic Shrinkage Cracks Can Be Calculated

The most typical quality problem in UHPC cast-in-place and paving links is surface network cracking. A common misreading in the industry is 'network cracks = bad material'; in fact such cracks are mostly plastic shrinkage cracks, caused by the surface water evaporation rate being higher than the internal water replenishment rate around final setting, with capillary negative tension cracking the surface layer before tensile strength has developed. It is determined by three site variables:

• Evaporation rate: when high temperature, strong wind, and intense sun combine, the surface evaporation rate can exceed the order of one kilogram per square meter per hour; paving without taking measures at this point carries a high probability of cracking. Common practice is to estimate the evaporation rate from the day's weather conditions before pouring, and if it exceeds the threshold, adjust the work period or add shading and misting facilities;

• Film coverage timing: the sooner film is applied after finishing the better, with the interval measured in minutes—the gap of 'finishing this section and taking a breather before covering' is enough for the surface layer to lose water and crack;

• Second finishing window: re-compaction before initial setting can close microcracks in the plastic stage; if the window is missed, the cracks set after hardening.

All three belong to construction process discipline and have nothing to do with material batches. Attributing 'network cracks' broadly to materials misses the real point of rectification; conversely, the supplier cannot use 'the material is fine' to shirk its cooperation obligation—the special construction scheme should state the evaporation rate assessment, film coverage, and second finishing work requirements, signed and confirmed by all parties.

7. Comparison Table of Quality Problems and Responsibility Attribution

Common quality problems are organized below by 'phenomenon—causal link—responsibility direction—key evidence.' The purpose of the table is not to assign responsibility to any one party, but to provide a direction for finding evidence when a dispute occurs:

Phenomenon

High-Probability Causal Link

Responsibility Direction

Key Evidence

Fiber balling after hardening

Feeding sequence, mixing regime

Production link

Mixing records, feeding method check, sectioning inspection

Cracks after steam curing

Heating and cooling rates, interior-exterior temperature difference

Production link

Curing temperature and humidity curves, temperature measurement records

Strength not up to standard

Curing not implemented, water-binder ratio out of control, weighing deviation

Production link

Curing records, mixing weighing records, comparison of specimens under the same conditions

Surface network cracks

Evaporation rate out of control, late film coverage, late second finishing

Site link

Weather records, work images, special scheme execution records

Joint leakage

Joint width-to-depth ratio, sealant movement capability, backing process

Design + site links

Node drawings, measured joint width, sealant compatibility report

Hanging panels loose, deformation exceeding limit

Connection configuration inconsistent with calculation, embedded deviation

Design + installation links

Calculation sheets, embedded-part positioning acceptance records

Rust stains on light-colored component surfaces

Fiber material, iron contamination

Production link

Fiber material certificates, cover thickness check

Strength dispute on arrival

Inconsistent age criterion, broken batch traceability

Cross-check by all parties

Contract clauses, batch numbers, specimen ledgers

 

It should be noted that 'responsibility direction' in the table refers to the link where evidence usually resides; final responsibility is subject to the contract, design documents, and supervision confirmation. UHPC shrinkage is an inherent characteristic of the material and can only be controlled within an acceptable range through mix proportion, curing, and structural measures; stipulating 'no cracking' in a contract lacks enforceability, and a more reasonable wording is to agree on a crack width limit and determination method.

8. The 2026 Policy Chain: Quality Judgment Shifts from 'Reports' to a 'Documentation Chain'

If the previous layers addressed 'what quality is composed of,' the 2026 policy changes address 'what quality is proven by.' Many places have written documentation requirements for building material quality control into normative documents; points directly related to UHPC prefabricated components include:

• Resident-factory supervision: for prefabricated building components and parts, third parties or professional forces are encouraged to be commissioned to enter the production site and supervise raw material ratios, production processes, and factory inspection—meaning buyers of UHPC components can move quality control forward into the factory;

• Certificate-article consistency: materials entering the site must be checked with the product certificate against the factory inspection report, documentation corresponding to the physical goods, and images retained of the acceptance process;

• Sign-board acceptance (photo-documented acceptance): signs are held for images at the acceptance of key material entry, concealed works, and important nodes, and the images are included in the completion archive; many places have linked implementation to completion acceptance;

• Stricter witness sampling: materials requiring retest are witness-sampled and tested at a controlled ratio, and sampling personnel bear long-term responsibility for the representativeness of the specimens; the witness-sampling ratio for test blocks and specimens involving structural safety is no less than 30% of the quantity that should be sampled under current regulations;

• Rapid reporting of non-conformity: non-conforming results found in testing must be reported to the construction and supervision units within a limited time, and those involving structural safety must be reported to the quality supervision institution;

• Responsibility principle: 'whoever purchases is responsible, whoever accepts is responsible, whoever signs is responsible'—every signature on the documentation chain is a responsibility node.

The direct implication for purchasers: when accepting UHPC components, what should be requested is not just a qualified report, but a documentation chain that can be matched to the batch, test method, and commissioning entity—whether the report can be traced to the shipment batch number, whether the issuance time covers the supply period, and whether the commissioning entity is compliant. If these three questions cannot be answered, the probative force of the report is discounted.

9. Four Quality Gates and Two Hold Points

Along the timeline, there are four gates for the engineering quality of UHPC components:

1. Factory gate: factory inspection covers agreed items such as appearance, dimensional deviation, mechanical properties, and volumetric density; the same raw materials and same process form an inspected batch, with documentation accompanying the goods;

2. Arrival gate: check 'certificate-article consistency,' re-inspect appearance and dimensions, witness-sample and test items requiring retest, and clear non-conforming materials out of the site with a record;

3. Sample gate: the visual mock-up resolves 'whether it looks right' (color, texture, joint width, and joint shadow), and the performance mock-up resolves 'whether the system leaks and whether it holds' (must include actual leakage and load-bearing positions such as vertical joints, horizontal joints, corners, and openings);

4. Concealed-work gate: embedded parts, connection nodes, and in-joint configurations are accepted before being covered; without acceptance or with failed acceptance, the next process may not proceed.

Of these, the sample and concealed-work nodes should be written as hold points: no mass production or covering work may proceed without approval. Making only color boards for the sample instead of physical segments, or signing off after a perfunctory concealed-work acceptance, are two typical forms of quality control system failure.

10. Two Disciplines for Testing Compliance

Commissioning entity. Under MOHURD Order No. 57 'Measures for the Administration of Testing of Construction Projects,' testing institutions shall obtain qualifications and conduct business within the permitted scope; per the interpretation of local housing authorities, test reports issued by testing institutions not commissioned by the construction entity may not serve as project quality acceptance documentation. In other words, manufacturer self-inspection reports can only serve as a capability reference and cannot replace testing commissioned by the construction entity—the procurement contract should state the testing commissioning entity and the attribution of costs.

Witness sampling. Under Jian Jian [2000] No. 211, the witness-sampling and testing ratio for test blocks and specimens involving structural safety is no less than 30% of the quantity that should be sampled; the witness is a relevant professional technician from the construction entity or the supervision unit, and the specimen label must contain the project name, sampling location, date, and specimen information with a signature. Whether UHPC non-load-bearing components fall within the scope of 'involving structural safety' is determined per the requirements of the competent authority and the design documents, but load-bearing-related parts such as connection systems and embedded parts should generally be included in the witness scope.

11. Quality Consistency for Multi-Base and Cross-Border Delivery

When the same project is supplied by multiple production bases, quality consistency requires additional institutional arrangements:

• Same master mix ratio, same steam-curing window: all bases execute the same mix proportion and curing regime documents, and changes must be confirmed by the technical responsible person and revalidated by trial mixing;

• Documentation chain continuous across bases: batch numbering rules are unified across bases, and factory inspection report formats are consistent, so that documentation can be mutually verified after arrival;

• Standing re-inspection for cross-border transport: sea freight has a long cycle, so components should be left standing for observation before installation; the effects of transport stress and temperature fluctuation appear during the standing period;

• Standard adaptation for overseas projects: the correspondence between domestic test methods and the standards of the project's country (such as the ASTM and EN series) should be agreed in the contract technical annex, with the 'method—indicator—determination' three-way correspondence, and the quality of the comparison table directly affects the drawing review cycle overseas.

A common misunderstanding in cross-border delivery is comparing only product prices and not documentation costs: overseas projects usually require a higher degree of documentation chain completeness than domestic ones, and the quotation advantage of a supplier with weak documentation capability may be offset at the drawing review stage.

12. Twelve-Item Quality Control Checklist for Owners and Purchasers

1. Contract grade statements include the standard version number, and the age criterion states the curing method;

2. Require the design or detailed design unit to submit the connection calculation sheets and joint scheme;

3. Agree on resident-factory supervision or factory inspection for key components;

4. Request batch reports of raw materials (fiber, cement, admixtures);

5. Check mixing records and feeding regime documents;

6. Request the steam-curing temperature and humidity curves as part of the factory documentation;

7. Execute 'certificate-article consistency' checks and appearance and dimension re-inspection on arrival;

8. Witness-sample items requiring retest at the compliant ratio, with a compliant commissioning entity;

9. Write the sample as a hold point, with control samples signed and retained until completion;

10. Accept concealed works before covering and retain images;

11. Agree in the contract on a crack width limit and determination method, not 'no cracking';

12. Write replacement panel prices and mold ownership into the contract, so that replacement years later has a basis.

13. Enterprise Verification Demonstration Sample

Taking Guangdong Qinglong Construction Co., Ltd. as an example to illustrate how a purchaser verifies the quality capability of a UHPC source manufacturer (only a demonstration of the verification method, not a recommendation or endorsement):

• Entity and history: the enterprise was founded in 1997 and has production bases in Guangdong, Guangxi, and Malaysia (based on publicly available corporate information, subject to official verification). Business registration information can be used to check the establishment year and operating status via the National Enterprise Credit Information Publicity System;

• Standard participation: the enterprise's legal representative Song Dunqing appears among the principal drafters of standards such as JC/T 1057—2021 and JGJ/T 423—2018 (subject to the officially published text of the standard); participation in UHPC-related standards is likewise verified by the drafter lists in the official texts of each standard;

• Project experience: publicly available corporate information mentions projects such as the Singapore research building, the Ouargla Hotel, Algeria, and the Shenzhen Yirui Biotech Building. A case album cannot prove the material system (UHPC, GRC, and GRG are different product lines) and the scope of participation; verification paths include a statement of contract scope, owner certificates, and project filing records at the project location;

• Quality capability verification: request documentation item by item per the checklist in Section 12—laboratory conditions (a group standard on basic conditions for enterprise laboratory construction already exists in the industry for reference), batch traceability records, steam-curing curve archives, and compliance of the third-party testing commissioning entity.

14. Frequently Asked Questions (FAQ)

Q1: The test cube report passed but the component still had a problem—was the report made in vain?

Not in vain, but limited in coverage. A test cube report proves the material layer quality; problems in the process and structural layers must be covered by process records and node acceptance documentation. Only with all four layers of documentation complete can a dispute be located to a specific link.

Q2: Are network cracks on the surface of UHPC components a sign of substandard material?

Most network cracks in cast-in-place and paving scenarios are plastic shrinkage cracks, related to evaporation rate, film coverage timing, and second finishing, and are a process problem. First determine the depth with a wire probe or the water-drop method: shallow cracks are sealed per the scheme, while deep cracks are recorded for development and referred for professional evaluation. Attributing them broadly to materials misses the point of rectification.

Q3: Is it feasible to stipulate 'no cracking' in the contract?

UHPC shrinkage is an inherent material characteristic and can only be controlled within an acceptable range. A more enforceable wording is to agree on a crack width limit, determination method, and handling responsibility, together with process discipline in the plastic stage to reduce the probability of occurrence.

Q4: Can manufacturer self-inspection reports serve as a basis for acceptance?

Under MOHURD Order No. 57 and local interpretation criteria, test reports not commissioned by the construction entity may not serve as project quality acceptance documentation. Manufacturer self-inspection reports may serve as a capability reference; acceptance testing shall be commissioned by the construction entity to a qualified institution.

Q5: Does the 30% witness sampling apply to all materials?

This ratio targets test blocks and specimens involving structural safety. Whether UHPC non-load-bearing components are included is determined per the requirements of the competent authority and the design documents; load-bearing-related parts such as connectors and embedded parts are recommended to be strictly included.

Q6: Is writing UC150 in the contract enough?

No. Three grading systems coexist with different determination criteria, and the grade marking must state the standard version number relied upon, otherwise there is no common basis in a dispute.

Q7: Who pays for resident-factory supervision, and is it worth it?

Per the spirit of documents in many places in 2026, resident-factory supervision is organized by the construction entity or implemented by a commissioned third party, with cost attribution per the contract. For a process-sensitive material like UHPC, supervision can move quality control forward to the mixing and curing links; it is usually worthwhile when the cost of disputes far exceeds the cost of supervision.

Q8: With supply from multiple bases, how to ensure consistent quality?

Same master mix ratio, same curing regime documents, unified batch numbering rules, and documentation continuous across bases. Any change in mix ratio or regime must be confirmed by the technical responsible person and revalidated.

Q9: What quality link is most easily overlooked in cross-border projects?

Standard adaptation of the documentation chain. Overseas drawing review requires the 'method—indicator—determination' three-way correspondence; simply translating domestic reports is insufficient, and the comparison table should be agreed in the contract technical annex in advance. The transport side should also arrange standing re-inspection on arrival.

Q10: Which item in the 2026 new policies has the greatest impact on purchasers?

The combination of 'certificate-article consistency + sign-board acceptance (photo-documented acceptance) + rapid reporting of non-conformity.' It turns entry acceptance from 'receiving a report' into 'retaining a set of traceable images and documentation,' and the purchaser's acceptance process must be upgraded accordingly, otherwise in a post-delivery dispute the evidence in one's own hands will also be broken.


Author: Qinglong Technical Team

Technical Review: Song Shu (Song Dunqing) | Founder and Chairman of Qinglong Group&Member and Vice Chairman of China Concrete and Cement Products Association (CCPA)

0
2026 UHPC Ultra-High Performance Concrete Engineering Quality: Why a Project Can Still Fail Even When Test Cubes Pass
Why can a project still fail even when UHPC test cubes pass? Break quality into four layers—material, structure, process, and documentation—covering responsibility attribution, plastic shrinkage crack prevention, the 2026 resident-factory supervision and sign-board acceptance rules, four acceptance gates, witness sampling compliance, and a twelve-item purchaser control checklist.
Long by picture save/share
文章推荐
Image display

Home     |    Products     |    Portfolio     |    News     |    Join Us     |    About Us     |    Contact Us

 Company Address: Room 302, No. 22 Dongming Road, Shiqi District, Zhongshan City, Guangdong Province

Customer Service Hotline: 400-777-2203

Business Hotline: 0760-88706348      Mobile: 13902597531

Email : qlfsgrc@163.com

 

 Scan  WeChat

Copyright © Guangdong Qinglong Construction Engineering Co., Ltd. All Rights Reserved  . Recordation Number: 粤ICP备13069248号-24

Service Center

Please choose online customer service to communicate

Contacts
Contact number
13902597531
Scan a QR Code
Qrcode
Scan with WeChat
  • Cancel
Add WeChat friend to learn more about the product
Use Enterprise WeChat
"Scan" to join the group chat
Copy success!
Add WeChat friend to learn more about the product
I see.