Material Testing and Quality Control
Material Properties Manxing Engineering Team 10 min read 2026-08-19
Material Properties Manxing Engineering Team 10 min read 2026-08-19

Material Testing and Quality Control in Industrial Silo Construction: What Actually Matters

Material testing and quality control in silo construction means verifying that every steel plate, weld, and bolt meets specification before it goes into a structure holding 10,000 tonnes of clinker. The two things that matter most? Mill certificates that trace back to the heat number, and weld inspection that catches defects before the silo is pressurized. Get those right and you've eliminated 80% of the failure risk.

Steel Plate Testing: Mill Certificates Are Not Optional

Material Testing and Quality Control

What a Mill Certificate Actually Tells You

A mill test certificate (MTC) under EN 10204 Type 3.1 is your baseline document. It lists the chemical composition, mechanical properties, heat number, and rolling mill of origin for every plate batch. I've seen projects in Indonesia where the MTC listed yield strength at 355 MPa, but spot-checking with a portable spectrometer showed actual values closer to 275 MPa. That's not a rounding error — that's a structural liability.

The certificate should specify carbon content (typically ≤0.20% for S355 grade), manganese, sulfur, phosphorus, and silicon percentages. For silo shells operating under cyclic loading from pneumatic conveying, you want Charpy V-notch impact values of at least 27 joules at -20°C. If the MTC doesn't include impact testing data, ask why. Some mills skip it to save $15 per test. That's a false economy when you're building a 30-meter-tall shell.

When to Re-Test Incoming Material

Don't trust the certificate blindly. We re-test every fifth heat number on projects above 5,000-tonne capacity. The protocol: portable XRF spectrometry for chemical verification, plus ultrasonic thickness measurement on 10% of plates. Tolerance on plate thickness is ±0.3 mm for plates under 12 mm, ±0.5 mm above that per EN 10029 Class A.

Here's a situation that still bothers me. A project in Vietnam — 2019, 8,000-tonne cement silo. The supplier delivered plates that looked right, certificates looked right. But the ultrasonic gauge showed three plates at 7.2 mm where 8.0 mm was specified. That's a 10% reduction in shell thickness. Over a 22-meter shell height, that translates to roughly 14% less buckling resistance under wind load. We rejected the entire batch. The supplier argued. We held firm. The replacement plates arrived two weeks late, but the silo's still standing.

Storage and Handling Before Fabrication

Plates arrive at the fabrication yard and immediately start corroding if you're within 5 km of the coast. I've measured 0.1 mm of surface rust on S355 plates after 72 hours of outdoor storage in Fujian province. That's not catastrophic, but it means you need abrasive blasting to SA 2.5 before coating — adding $8-12 per square meter to the surface prep budget.

Stack plates on timber bearers, not directly on ground. Keep them covered with breathable tarps — not plastic sheeting, which traps condensation. Separate grades by color-coding the heat number tags. Mixing S235 and S355 plates in the same shell course is a mistake I've seen twice. Both times, the fabricator caught it during layout. Both times, it was a near miss.

Weld Quality Control: The Inspection Hierarchy

Visual Inspection Is Your First Line — Not Your Last

Every weld gets a visual inspection per ISO 5817 Level B (stringent) for silo shells. That means no cracks, no lack of fusion, undercut ≤0.5 mm, and reinforcement within specified limits. A good inspector with a weld gauge and a flashlight catches 60-70% of surface defects in the first pass.

But visual inspection misses what's underneath. I watched a radiographic test in Shandong reveal a 40-mm lack of fusion in a circumferential seam that looked perfect from the outside. The welder had run the first pass too fast — travel speed was 350 mm/min when the WPS specified 200-250 mm/min. That single defect would've been a crack initiation point under cyclic pressure loading. We cut it out and re-welded. Cost us one day. Saved us a potential failure in year five.

Non-Destructive Testing: RT vs. UT vs. MT

For butt welds in the shell, radiographic testing (RT) per ISO 17636 gives you the best record — you get a permanent film or digital image. But RT is slow. A single 10 mm thick, 300 mm long weld takes 15-20 minutes including setup. On a silo with 4,000 linear meters of shell welds, doing 100% RT would add three weeks to the schedule.

The practical approach: 10% RT on circumferential seams, 5% on longitudinal seams, plus 100% ultrasonic testing (UT) per ISO 17640 on all Category 1 joints. Magnetic particle testing (MT) per ISO 17638 on all fillet welds connecting stiffeners to the shell. That's the inspection matrix we use on projects following EN 1993-4-1 (Eurocode 3, Part 4-1 for silos).

UT is faster than RT — a skilled technician can inspect 8-10 meters of weld per hour. But UT requires calibration on reference blocks with side-drilled holes. If the technician doesn't recalibrate every 4 hours or after every probe change, you get false negatives. I've seen it happen. The data looks clean. The weld isn't.

Welding Procedure and Welder Qualification

Before a single arc strikes on the project, you need a qualified welding procedure specification (WPS) per ISO 15614-1. The procedure qualification record (PQR) must cover the actual material grade, thickness range, and position that will be used on the silo. A PQR qualified for 12 mm plate in the flat position doesn't automatically cover 18 mm in the vertical position.

Every welder on the project needs a current qualification certificate per ISO 9606-1, valid for the specific process (111 for MMA, 135 for MAG, 141 for TIG) and joint type. Certificates are valid for six months with continuous employment, or three months if the welder hasn't been working. I've turned away qualified welders whose certificates had lapsed by two weeks. The client didn't understand. I explained that if a weld fails and the insurance investigator finds an expired welder cert, the claim gets denied. They understood that.

Bolted Connection Quality: Torque, Tension, and the Gap Between Them

Why Torque Wrenches Lie to You

A torque wrench tells you the torque applied to a nut. It does not tell you the actual clamp force in the bolt. The relationship between torque and tension depends on the friction coefficient, which varies by ±30% depending on surface condition, lubrication, and whether the bolt has been previously tightened.

For high-strength structural bolting per EN 14399 (preloaded bolting assemblies), the combined method is the standard approach: tighten to the snug-turn stage, then apply the specified rotation (typically 120°-180° for Grade 10.9 bolts). The turn-of-nut method is more reliable than torque alone because it's based on bolt elongation, not friction.

We specify a minimum preload of 70% of the bolt's proof load for connections subject to cyclic loading. For an M24 Grade 10.9 bolt, that's approximately 225 kN. Achieving that consistently requires calibrated equipment and a crew that understands the procedure. I've seen crews hit 180 kN on the same bolt specification because they were using a torque wrench that hadn't been calibrated in eight months.

Gasket and Sealant Verification

For cement and clinker silos with bolted shell joints, the gasket material matters more than most engineers realize. EPDM gaskets rated for continuous service at 120°C are standard. But if you're storing hot clinker at 200-300°C, you need silicone or fluoropolymer gaskets. I've seen EPDM gaskets on a slag silo harden and crack within 18 months because the material was rated for 120°C and the actual service temperature hit 160°C during discharge cycles.

Sealant application on flanged roof joints and wall penetrations needs verification too. We pull-test sealant beads after curing — minimum adhesion of 0.8 MPa on steel substrate per ISO 8339. If the bead peels off with finger pressure, it's not sealing anything.

Concrete Foundation and Anchor Bolt Quality

Anchor Bolt Positioning: Millimeters Matter

Anchor bolt tolerance for silo foundations is ±3 mm in position, ±1 mm in projection height. That sounds tight until you realize the steel shell base ring has holes drilled at ±2 mm tolerance. Stack those tolerances and you've got a potential 5 mm mismatch. On a 16-meter diameter silo with 48 anchor bolts, that's manageable. On a 30-meter silo with 96 bolts, it's a geometry problem.

We use steel templates — not wooden jigs — to hold anchor bolts in position during concrete pour. The template stays in place for 24 hours after pour, then gets removed and checked with a total station. I've measured anchor bolt positions at 3am under floodlights because the concrete truck was arriving at 6am and we needed to verify before the pour. The crew wasn't happy. The bolts were within spec. That's what mattered.

Concrete Strength Verification

Foundation concrete for large silos typically specifies C35/45 (35 MPa cylinder strength, 45 MPa cube strength). We cast 150 mm cube specimens at the pour site — minimum 6 cubes per 100 m³ of concrete — and test at 7 days and 28 days per EN 12390. The 7-day result should reach at least 65% of the 28-day characteristic strength. If it doesn't, something's wrong with the mix or the curing.

On a project in Myanmar, the 7-day cubes came back at 18 MPa when we expected 24 MPa. Ambient temperature had been 38°C with low humidity — the concrete was curing too fast on the surface but not gaining strength in the core. We switched to wet burlap curing with plastic sheeting and added a set retarder to the mix. The 28-day results hit 41 MPa. Problem solved, but it cost us four days of schedule.

Coating and Corrosion Protection Quality

Surface Preparation Standards

The coating is only as good as the surface it's on. For silo exteriors in industrial environments, abrasive blasting to SA 2.5 per ISO 8501-1 is the minimum. That means the surface is free from visible oil, grease, and mill scale, with any remaining contamination limited to randomly distributed light stains.

Surface profile after blasting should be 50-75 microns (Rz) for most epoxy systems. We measure it with replica tape per ISO 8503-5. Too smooth and the coating won't adhere. Too rough and you get peak-to-valley holidays where the coating is thinner than specified. I've seen profiles hit 120 microns on a coastal project in Gujarat — the blaster was using oversized grit. The coating system was specified at 250 microns DFT, but the peaks were sticking through at 180 microns. We had to apply an extra stripe coat. Added $6/m² to the coating cost.

Dry Film Thickness Measurement

Measure DFT per ISO 2808, Method 6A, using a magnetic or eddy current gauge. Take 5 readings per square meter, average them, and ensure 90% of readings meet or exceed the specified minimum. No single reading should fall below 80% of the specified DFT.

For a typical silo coating system — zinc-rich epoxy primer at 75 microns, epoxy intermediate at 150 microns, polyurethane finish at 50 microns — the total system is 275 microns DFT. That's the spec for a C3 corrosivity category per ISO 12944. If the silo is within 2 km of the coast, you're in C4 or C5 territory and the system jumps to 320-350 microns. Budget accordingly.

Field Tip: Always calibrate your DFT gauge on a smooth reference surface before each measurement session, then verify on a certified foil of known thickness. A gauge that's off by 20 microns on a 250-micron system is an 8% error — enough to reject a coating that actually passes, or accept one that doesn't. I've seen both. Calibrate every morning. Takes 90 seconds.

Case Study: 12,000-Tonne Clinker Silo — Quality Control Failure and Recovery

A cement plant in East Africa commissioned a 12,000-tonne clinker silo in 2021. The EPC contractor (not Manxing) had subcontracted the shell fabrication to a local shop with limited QA infrastructure. During erection, our team was called in for a third-party inspection after the client noticed visible waviness in the shell plates.

The findings were systematic. First, the shell plates showed thickness variations of ±0.8 mm on nominal 10 mm plates — outside EN 10029 Class A tolerance. Second, UT inspection of 20% of circumferential seams revealed 14% had lack of fusion defects exceeding ISO 5817 Level B acceptance criteria. Third, anchor bolt positions were off by up to 12 mm — four times the specified tolerance. The base ring had been forced into position, inducing residual stress in the shell.

The recovery plan: reject and replace 30% of shell plates, re-weld 220 meters of defective seams, and re-drill the base ring holes with a portable CNC machine to match actual anchor positions. Total cost of remediation was $340,000 and a 10-week schedule delay. The original fabrication cost had been $1.2 million. The QA/QC budget on the original contract was $18,000 — 1.5% of fabrication cost. Industry best practice is 4-6%. That gap paid for itself in reverse.

FAQ

Q: How often should weld inspection be performed during silo erection?

Inspection happens in three stages: fit-up (before welding), in-process (during multi-pass welds, checking interpass temperature), and post-weld (visual, then NDT). For a typical 10,000-tonne silo with 3,500 meters of shell welds, plan for 200-250 visual inspections and 150-180 NDT examinations. That's roughly one inspection point per 14 meters of weld. Don't batch all NDT to the end — you'll find defects in welers that were done in week two, and the crew that made them has moved to another site.

Q: What's the acceptable rejection rate for incoming steel plates?

Under 2% is normal. Above 5% and you have a supplier quality problem, not a random variation issue. We track rejection rate by heat number and by supplier. If a supplier hits 5% rejection on two consecutive shipments, we switch to 100% incoming inspection on their material and notify them in writing. Most suppliers respond within one shipment. The ones that don't get removed from our approved vendor list.

Q: Can you use the same coating system for interior and exterior silo surfaces?

Usually not. Exterior surfaces need UV resistance — that's where the polyurethane finish coat comes in. Interior surfaces in a cement silo need abrasion resistance from material flow, and chemical resistance from alkaline dust (pH 11-12 for cement). We specify epoxy novac interior linings at 400-500 microns DFT for cement and clinker service. For coal silos, the interior needs static conductivity — surface resistance below 10⁹ ohms per IEC 60079-32-1 to prevent electrostatic discharge. Different environments, different systems. Don't let a painter talk you into "just using the same stuff inside and out."

Need a Quality Control Plan for Your Silo Project?

Manxing provides full EPC silo contracting with integrated material testing and quality control — from mill certificate verification through final coating inspection. Our QA/QC protocols follow EN 1090, ISO 3834, and project-specific standards across 30+ countries.

Contact Manxing for a Project Assessment →

Written and reviewed by the Manxing silo engineering team · Last updated Oct 9, 2026

Prepared by our in-house bulk-handling engineers and checked against the design practice listed below. Questions about a specific project can be sent to our engineering desk through the contact page.

Standards and industry references

  • EN 1991-4 — Eurocode 1: Actions on structures, Part 4: Silos and tanks
  • ACI 313 — Design and construction of concrete silos and stacking tubes
  • GB 50077 — Code for design of reinforced concrete silos
  • ISO / ASABE bulk-solids handling guidance on material flow and segregation
Need Engineering Support?
Talk to our engineers about your specific requirements
Talk to an Engineer
Tel: +86 159 3903 7000
[email protected]
+86 159 3903 7000