Non-Destructive Evaluation and Structural Health Monitoring

Introduction and Context

Non-destructive evaluation (NDE) comprises methods used to investigate the condition, uniformity, defects and in-situ properties of concrete without significantly damaging the structure. Structural health monitoring (SHM) extends this idea by using repeated measurements—often from permanently installed sensors—to observe structural response, detect change and support maintenance decisions.

Concrete structures are affected by construction variability, cracking, inadequate compaction, reinforcement corrosion, chemical attack, fatigue, overload, fire and environmental exposure. Visual inspection remains essential, but it cannot reliably identify hidden voids, internal cracking, debonding, loss of reinforcement section or changes in stiffness. NDE and SHM therefore provide complementary evidence for condition assessment.

A central principle is that NDE measurements are usually indirect. A rebound number, ultrasonic pulse velocity or half-cell potential is not itself a direct measurement of compressive strength, crack severity or corrosion rate. Reliable assessment requires calibration, knowledge of construction history, environmental correction, multiple techniques and engineering interpretation.

Learning Objectives

After completing this chapter, the reader should be able to:

1. Explain the principles and applicability of major NDE methods for concrete.

2. Distinguish screening, diagnosis, quantification and monitoring tasks.

3. Plan an investigation, including sampling, calibration, quality control and uncertainty assessment.

4. Interpret rebound, ultrasonic, impact-echo, electromagnetic, acoustic, thermal and sensor-based data.

5. Combine independent measurements to support decisions on repair, strengthening or continued service.

6. Recognise limitations, ethical responsibilities and emerging digital approaches in SHM.

1. NDE and SHM as an Engineering Decision Process

A professional investigation normally follows six stages:

Define the question. Examples include: Is the concrete sufficiently uniform? Is a suspected delamination present? How extensive is corrosion? Has a strengthening intervention improved performance?

Develop a conceptual model. Review drawings, materials, construction records, exposure conditions, loading, previous repairs and visible symptoms. The same measurement can have different meanings in different structural systems.

Select complementary methods. A rapid method may screen large areas, while a more local method confirms the suspected mechanism. Limited destructive tests, such as cores or reinforcement exposure, are often necessary for calibration and validation.

Acquire and quality-check data. Record instrument settings, locations, surface condition, temperature, moisture, operator and environmental conditions. Measurements should be spatially referenced.

Interpret uncertainty. Results should be expressed with confidence limits or decision ranges where possible, not as unjustifiably precise single values.

Decide and verify. Findings may lead to monitoring, additional testing, repair, strengthening, load restriction or demolition. Post-intervention testing verifies effectiveness.

NDE is commonly classified as: (i) visual and dimensional inspection, (ii) mechanical surface-response methods, (iii) wave-propagation methods, (iv) electromagnetic and electrochemical methods, (v) thermal and acoustic methods, and (vi) permanently or semi-permanently installed sensor systems.

2. Rebound Hammer Testing

Principle

A spring-driven mass impacts a plunger held against concrete. The rebound distance is expressed as a rebound number. Harder surfaces generally produce higher rebound numbers because they absorb less impact energy.

Application and procedure

The surface should be clean, sound and free of loose laitance, coatings and standing water. Readings are taken at multiple nearby locations using consistent orientation. Tests should avoid aggregate particles, edges, honeycombing and reinforcement located close to the surface. The instrument must be checked against its reference anvil according to the manufacturer’s procedure.

The rebound number is influenced by surface moisture, carbonation, age, aggregate type, impact direction, surface roughness, temperature and operator technique. Carbonated surfaces may give an artificially high result. A strength estimate should therefore be based on a structure-specific correlation established using cores or representative test specimens whenever safety-critical decisions are involved.

A common statistical model is a regression such as

\[

f_c=a+bR,

\]

where \(f_c\) is estimated compressive strength, \(R\) is rebound number, and \(a\) and \(b\) are empirically determined coefficients. This is not a universal law. The equation is applicable only to concrete comparable with the calibration population and within its measured range. A nonlinear regression may be more appropriate for some datasets. Reporting should include the calibration method, coefficient of determination, prediction error and range of validity.

The rebound hammer is particularly useful for mapping uniformity and identifying anomalously soft or hard zones. It is weak as a stand-alone method for determining absolute strength.

3. Ultrasonic Pulse Velocity

Principle and measurement

Ultrasonic pulse velocity (UPV) measures the travel time of a stress wave between a transmitting and receiving transducer. For a direct transmission path,

\[

V=\frac{L}{t},

\]

where \(V\) is pulse velocity, \(L\) is the path length and \(t\) is the measured travel time. If \(L\) is in metres and \(t\) in seconds, \(V\) is in metres per second.

Direct transmission through opposite faces generally gives the most reliable result. Semi-direct and indirect arrangements are useful where access is limited but require careful interpretation. Coupling material, transducer pressure, surface roughness, path geometry and equipment settings affect results.

Higher velocity often indicates greater uniformity and continuity, but velocity is affected by aggregate mineralogy, moisture, age, temperature, stress state, reinforcement and mix proportions. Cracks, voids and delamination increase travel time or prevent transmission. Steel reinforcement may create a high-velocity path and cause overestimation if its location and orientation are ignored.

UPV is excellent for locating discontinuities and comparing zones. Strength estimation requires a project-specific correlation because the relationship between velocity and compressive strength is not unique. Tomographic imaging, using measurements over many paths, can provide a two- or three-dimensional representation of internal anomalies, although resolution depends on wavelength, geometry and data density.

4. Combined Rebound–Ultrasonic Assessment

Combining rebound number and UPV can reduce some individual-method limitations. A generic multivariable model may be written as

\[

f_c=aR^bV^c,

\]

where \(f_c\) is estimated strength, \(R\) is rebound number, \(V\) is pulse velocity, and \(a\), \(b\), and \(c\) are calibration constants. The exponents and coefficients must be obtained from project-relevant paired measurements; they must not be borrowed indiscriminately from another structure.

The combined method can distinguish, for example, a dry carbonated surface (high rebound but not necessarily high bulk quality) from genuinely dense concrete with high velocity. Nevertheless, correlated errors, inappropriate calibration and differing measurement locations can create false confidence. Statistical validation and residual analysis are essential.

5. Impact-Echo and Related Wave Methods

Impact-echo testing uses a short mechanical impact to generate stress waves. Reflections from interfaces, cracks, voids and the opposite surface are detected by a receiver. Frequency-domain analysis identifies dominant resonant frequencies. For a simple plate-like member, thickness may be estimated from

\[

d=\frac{C_p}{2f},

\]

where \(d\) is thickness, \(C_p\) is the relevant compressional-wave velocity and \(f\) is the dominant thickness resonance frequency. The equation assumes a relatively uniform plate, a suitable mode identification and known or independently estimated wave velocity. In practice, geometry, reinforcement, multiple reflections and mode conversion complicate interpretation.

Impact-echo is useful for estimating slab thickness and locating delamination, voids and debonding. It can investigate one-sided access conditions. Surface-wave methods, including spectral analysis of surface waves, estimate near-surface stiffness profiles from dispersion. These methods require skilled interpretation because surface finish, geometry and nearby boundaries influence the response.

6. Electromagnetic and Electrochemical Methods

Reinforcement location and cover

Cover meters and electromagnetic scanners detect changes in electromagnetic fields caused by steel reinforcement. They can map bar spacing, orientation and approximate concrete cover. Results are affected by bar diameter, adjacent bars, congested reinforcement, metallic formwork and the instrument’s calibration. Verification with drawings or limited exposure is prudent.

Corrosion potential

Half-cell potential measurements compare the electrical potential of reinforcing steel with a reference electrode. They indicate the likelihood of active corrosion under suitable electrical continuity and moisture conditions. Potential values are not direct corrosion rates and cannot alone quantify metal loss. Interpretation requires consideration of concrete resistivity, carbonation, chloride exposure, cracking, moisture and electrical isolation.

Concrete resistivity

Resistivity measurements indicate the ease with which ionic current can pass through concrete. Lower resistivity generally supports a greater likelihood of corrosion activity when other conditions are present, while high resistivity tends to limit corrosion current. However, resistivity is strongly moisture- and temperature-dependent. It should be used with chloride/carbonation information, potential mapping and visual evidence.

Ground-penetrating radar

GPR sends electromagnetic pulses into concrete and records reflections from reinforcement, interfaces, voids and moisture changes. It is effective for rapid scanning of large areas and estimating cover or identifying anomalies. High-frequency antennas provide better resolution but less penetration. Interpretation is affected by conductive concrete, moisture, dense reinforcement and signal attenuation.

7. Acoustic Emission Monitoring

Acoustic emission (AE) detects transient elastic waves generated by active processes such as crack growth, friction, reinforcement slip and debonding. Sensors attached to the surface record arrival time, amplitude, energy, duration and frequency characteristics.

A cluster of AE events may indicate a developing damage zone, while source-location algorithms can estimate the position of activity using differences in arrival time. Accurate localisation requires sensor calibration, known wave velocity and adequate sensor geometry. Background noise from traffic, machinery, rain and electrical sources must be filtered or identified.

AE is especially valuable during load tests because it can reveal damage activity while the structure is being stressed. It is less suitable for identifying old, inactive cracks, which may produce no current emissions. AE should therefore complement visual, geometric and other NDE observations.

8. Infrared Thermography

Infrared thermography records surface temperature patterns. Defects such as delamination, voids or debonding alter heat flow and thermal capacity, producing temperature contrasts under suitable heating or cooling conditions. Passive thermography uses natural environmental heating; active thermography uses a controlled heat source.

The method is rapid and non-contact, but results depend on solar radiation, wind, surface emissivity, moisture, coatings, surface geometry and the depth and size of the defect. A thermal image is an anomaly map, not a direct defect diagnosis. Suspected areas should be confirmed by another method, such as impact-echo, sounding or local opening.

9. Sensor-Based Structural Health Monitoring

SHM systems combine sensors, data acquisition, communications, analysis and decision rules. Typical measurements include strain, acceleration, displacement, inclination, crack opening, temperature, humidity, corrosion indicators, acoustic emission and load.

Monitoring architectures

A periodic system uses portable instruments during scheduled inspections. A continuous system records data automatically, often with event-triggered or scheduled sampling. A hybrid system combines continuous low-rate measurements with high-rate recording during unusual events.

Baselines and thresholds

A baseline should be established under known structural and environmental conditions. Raw measurements commonly contain environmental and operational effects. A simple decomposition is

\[

y(t)=s(t)+e(t)+n(t),

\]

where \(y(t)\) is observed response, \(s(t)\) is structural response, \(e(t)\) is environmental or operational influence, and \(n(t)\) is measurement noise. The terms are conceptual; separating them requires calibration, environmental sensors and statistical modelling.

Thresholds may be based on engineering limits, control charts, model predictions or changes from a validated baseline. A threshold should specify the action that follows: confirmation, increased sampling, inspection, load restriction or emergency response. Alarm systems without an action protocol create either complacency or alarm fatigue.

Data quality

Sensor drift, loss of power, time synchronisation errors, communication failure and poor installation can produce misleading trends. Quality assurance should include redundancy where justified, calibration checks, automated range checks, missing-data flags, audit trails and periodic physical inspection.

10. Integrated Condition Assessment

No single NDE method answers every structural question. An integrated matrix can link symptoms to methods:

• Low strength or poor uniformity: rebound, UPV, cores and construction records.
• Internal voids or honeycombing: UPV, impact-echo, GPR and selective opening.
• Delamination: impact-echo, thermography, chain drag or sounding.
• Reinforcement cover and layout: electromagnetic scanning and drawings.
• Corrosion likelihood: half-cell potential, resistivity, chloride/carbonation testing and visual inspection.
• Active cracking or debonding: AE, crack gauges, vibration and repeated imaging.
• Excessive movement or loss of stiffness: displacement, strain, vibration and load testing.

Bayesian reasoning can be useful: an initial hypothesis about a defect is updated as independent evidence is obtained. Independence should not be assumed automatically; rebound and UPV, for example, may both be affected by moisture and age. The final report should distinguish observation, interpretation, inferred cause and confidence level.

11. Worked Example: Estimating Pulse Velocity and Interpreting Uniformity

A direct UPV test is performed across a concrete wall. The measured path length is 0.600 m and the travel time is 150 microseconds.

Convert time:

\[

150\ \mu s=150\times10^{-6}\ s=0.000150\ s.

\]

Therefore,

\[

V=\frac{0.600}{0.000150}=4000\ m/s.

\]

The result is a measured pulse velocity of 4000 m/s. It should not be converted directly into compressive strength without a validated project-specific relationship. Suppose repeated measurements in adjacent locations are 3980, 4010, 4000, 3150 and 3970 m/s. The isolated low value is a diagnostic anomaly, not automatically proof of a void. The location should be retested, coupling checked, reinforcement geometry reviewed and, if the anomaly persists, examined using impact-echo, GPR or a core. This illustrates why spatial patterns and confirmation are more informative than a single number.

12. Illustrative Case Study: Parking-Garage Deck

Background. An eight-year-old reinforced-concrete parking deck in a coastal environment exhibits rust staining and isolated delamination near expansion joints.

Problem. The owner must decide whether local patch repair is adequate or whether corrosion is widespread.

Analysis. A visual grid survey maps staining, cracking and delamination. Cover-meter scanning identifies shallow cover near several joints. Half-cell mapping indicates areas with more negative potentials, while resistivity is low in damp zones. Thermography and impact-echo identify delaminated regions. Chloride samples at selected depths confirm that chloride concentration is elevated near the reinforcement in representative locations.

Approach. The investigation team calibrates the survey with limited openings and cores. Delaminated concrete is removed, reinforcement loss is measured, chloride-contaminated concrete is assessed, and repair boundaries are extended beyond visibly loose concrete where testing supports continuation of corrosion risk. Drainage and joint sealing are improved. Sensors measuring deck moisture and corrosion-related electrochemical response are installed at representative high-risk locations.

Outcome. Local repairs are combined with waterproofing and improved drainage rather than treating staining alone. Follow-up potential, resistivity and visual surveys are scheduled after wet and dry seasons. The monitoring system is linked to an inspection protocol rather than used as an autonomous repair decision.

Learning points. Corrosion diagnosis requires multiple lines of evidence. Potential mapping identifies likelihood, not metal loss. Environmental control can be as important as patching. Calibration and follow-up are necessary to determine whether the intervention has changed the deterioration process.

13. Limitations, Ethics and Governance

NDE conclusions can affect public safety, insurance, legal responsibility and major expenditure. Practitioners must state competence limits, instrument limitations and uncertainty. It is unethical to present empirical correlations as universal material laws or to suppress inconvenient results. Data should be traceable to locations, equipment and operators.

Access, privacy and cybersecurity also matter in SHM. Cameras, drones and connected sensors may collect information about occupants, workers or facility operations. Data governance should define ownership, retention, access rights, cybersecurity controls and responsibilities for responding to alarms. Digital records must be preserved sufficiently to support future audits.

Regulatory compliance should follow the applicable national standards, project specifications and professional guidance for testing, structural assessment, load testing, electrical safety and work at height. NDE does not remove the requirement for engineering judgement or, where necessary, destructive verification.

14. Contemporary Developments and Best Practice

Modern developments include robotic scanners, unmanned aerial vehicles, distributed fibre-optic strain sensing, wireless sensor networks, digital twins, automated image analysis and machine-learning classifiers. These technologies can improve coverage and detect patterns, but they do not eliminate the need for representative training data, explainability, calibration and expert review.

A machine-learning model trained on one concrete type, camera system or environmental condition may fail elsewhere. Algorithms should be tested against labelled field data, uncertainty should be reported, and automated classifications should not be treated as final structural decisions without independent verification.

Best practice is to:

1. Begin with a failure-mode and exposure-based inspection plan.

2. Use calibrated instruments and document environmental conditions.

3. Combine rapid scanning with targeted confirmatory tests.

4. Establish spatially referenced datasets and repeatable measurement locations.

5. Separate screening results from quantified engineering parameters.

6. Report uncertainty, limitations and alternative explanations.

7. Define alarm actions before installing monitoring equipment.

8. Reassess after repair, unusual loading or significant environmental change.

9. Preserve raw data, metadata and an auditable decision trail.

Chapter Summary

NDE and SHM provide complementary means of diagnosing concrete deterioration, estimating selected in-situ properties and tracking structural response. Rebound testing is rapid but surface-sensitive; UPV evaluates continuity and uniformity; impact-echo investigates thickness and internal interfaces; electromagnetic methods locate reinforcement and support corrosion assessment; AE detects active damage; thermography identifies thermally distinguishable anomalies; and sensors provide repeated response data. None should be interpreted without regard to environmental effects, calibration and uncertainty. Reliable condition assessment integrates several methods with drawings, exposure history, visual inspection and selective destructive verification. SHM is most valuable when baseline behaviour, data quality, alarm actions and governance are clearly defined.

Key Terms

Non-destructive evaluation: Investigation that causes negligible or acceptable damage to the structure.

Structural health monitoring: Repeated measurement and interpretation of structural condition or response.

Rebound number: Instrument reading representing surface impact rebound.

Ultrasonic pulse velocity: Stress-wave speed measured through concrete.

Impact-echo: Wave method using impact-generated reflections to identify interfaces and defects.

Tomography: Imaging based on measurements along multiple paths.

Electromagnetic scanning: Detection of reinforcement and interfaces through electromagnetic response.

Half-cell potential: Electrochemical potential used to assess corrosion likelihood.

Concrete resistivity: Opposition of concrete to ionic current flow.

Ground-penetrating radar: Electromagnetic pulse method for subsurface imaging.

Acoustic emission: Transient elastic-wave activity produced by active damage processes.

Thermography: Imaging of surface temperature distribution.

Delamination: Separation of concrete layers, commonly parallel to a surface.

Calibration: Establishment of a measurement relationship using known or independently tested values.

Baseline: Validated reference behaviour used for comparison.

Sensor drift: Gradual change in sensor output unrelated to structural change.

Prediction uncertainty: Uncertainty associated with estimating an unmeasured property.

Digital twin: Data-linked digital representation of a physical asset.

Condition assessment: Engineering evaluation of current state, deterioration and serviceability.

Review Questions

Short Answer Questions

1. Why is a rebound hammer not normally sufficient for determining concrete strength?

2. State the equation for direct ultrasonic pulse velocity and define its variables.

3. What type of information does half-cell potential provide?

4. Why should thermal anomalies be confirmed by another method?

5. Distinguish periodic monitoring from continuous monitoring.

Long Answer Questions

1. Explain the principles, applications and limitations of rebound hammer and UPV testing.

2. Discuss how electromagnetic and electrochemical methods can be combined to assess reinforcement corrosion.

3. Describe a complete investigation plan for suspected delamination in a reinforced-concrete slab.

4. Explain the role of baseline modelling, environmental compensation and data quality in SHM.

5. Critically evaluate the use of machine learning in automated concrete defect detection.

Analytical / Application Questions

1. A UPV path is 0.75 m and the measured travel time is 187.5 microseconds. Calculate the velocity and explain what additional information is required before estimating strength.

2. A deck has high rebound numbers, low UPV in isolated zones and shallow reinforcement cover. Propose a testing sequence and justify each method.

3. Design an SHM and follow-up inspection strategy for a coastal parking structure showing rust staining and seasonal cracking.

Multiple Choice Questions with Answers

Question 1

What is the principal use of rebound hammer testing in a large concrete structure?

A. Direct measurement of reinforcement loss

B. Mapping relative surface uniformity

C. Measurement of chloride concentration

D. Determination of crack growth rate

Correct Answer: B. Mapping relative surface uniformity

Question 2

A UPV path is 0.600 m and travel time is 150 microseconds. What is the pulse velocity?

A. 400 m/s

B. 2,500 m/s

C. 4,000 m/s

D. 90,000 m/s

Correct Answer: C. 4,000 m/s

Question 3

Which factor can cause rebound testing to overestimate the strength of the concrete interior?

A. Surface carbonation

B. Low reinforcement cover only

C. High internal moisture only

D. Long ultrasonic path length

Correct Answer: A. Surface carbonation

Question 4

Half-cell potential measurements primarily indicate the:

A. Exact diameter loss of reinforcement

B. Likelihood of active corrosion

C. Compressive strength of concrete

D. Thickness of a concrete slab

Correct Answer: B. Likelihood of active corrosion

Question 5

Which method is particularly suitable for detecting active crack growth during a load test?

A. Acoustic emission

B. Cover-meter scanning

C. Passive thermography only

D. Rebound hammer testing

Correct Answer: A. Acoustic emission

Question 6

Impact-echo testing uses reflections of waves generated by:

A. A controlled mechanical impact

B. A chemical chloride reaction

C. A half-cell electrode

D. A thermal camera

Correct Answer: A. A controlled mechanical impact

Question 7

In the monitoring expression y(t)=s(t)+e(t)+n(t), e(t) represents:

A. Structural response

B. Environmental or operational influence

C. Sensor calibration constant

D. Concrete elastic modulus

Correct Answer: B. Environmental or operational influence

Question 8

Why should GPR results in heavily reinforced concrete be interpreted cautiously?

A. Electromagnetic signals may be attenuated and reflections may overlap

B. GPR measures only surface hardness

C. GPR cannot detect any metallic object

D. GPR directly measures compressive strength

Correct Answer: A. Electromagnetic signals may be attenuated and reflections may overlap

Question 9

Which statement about a project-specific strength correlation is correct?

A. It is universally valid for all aggregate types

B. It should be calibrated using relevant concrete data

C. It eliminates all measurement uncertainty

D. It can be used outside its calibration range without qualification

Correct Answer: B. It should be calibrated using relevant concrete data

Question 10

What is an essential feature of an SHM alarm system?

A. A threshold with no prescribed response

B. A response protocol linked to the alarm

C. Permanent removal of manual inspections

D. Use of a single uncalibrated sensor

Correct Answer: B. A response protocol linked to the alarm