UIC 513: Ride Quality Standards with Comfort Index Scale Explained
UIC 513 defines the NMV Comfort Index methodology for evaluating passenger ride quality in railway vehicles. Learn frequency weighting filters, sensor placement protocols, six-grade comfort scale, and how UIC 513 compares with EN 12299 and ISO 2631.
⚡ IN BRIEF
- Current edition: 1st edition (1 July 1994, 13 pages), with a revised edition published 12 January 2013 (status: current). The leaflet is available in English, German, and French. (Source: Normadoc; Standards.ie)
- Primary objective: To define test conditions, measurement parameters, and evaluation methods for quantifying passenger comfort in relation to vibration in railway vehicles, with the aim of reducing passenger discomfort through improved vehicle design and operation. (Source: renrendoc.com)
- Simplified and complete measurement methods: The leaflet permits two assessment approaches: a simplified method (using floor‑level accelerations only) and a complete method (using floor‑level and seat‑level accelerations). Both methods yield a comfort index, NMV. (Source: renrendoc.com)
- Quantitative comfort thresholds: The leaflet defines a six‑grade subjective scale based on the NMV index: <1 (very good), 1–2 (good), 2–4 (moderate), 4–5 (poor), >5 (very poor comfort). (Source: Academic OUP Preview)
- Reference for EN 12299 and superseded for EU new builds: The leaflet is the direct basis for the European standard EN 12299. For new rolling stock placed on the EU market, EN 12299 is the harmonised standard; however, UIC 513 remains applicable for non‑EU fleets and legacy vehicles. (Source: DIN; Korea Science PDF)
In 2005, a newly delivered fleet of intercity EMUs entered service on a cross‑border route connecting Germany, Austria and Switzerland. Within months, passenger complaints about motion sickness and discomfort on the mountainous sections of the line reached an unprecedented level. The operator commissioned an independent investigation, which measured vibration accelerations in the passenger saloon and calculated the UIC 513 comfort index. The result was an NMV value of 3.2 — “moderate comfort” — substantially worse than the contractual target of ≤ 2.0 (“good comfort”). The manufacturer had relied on computer simulations using generic track excitation inputs rather than performing the on‑track measurements required by the leaflet. The fleet was recalled for retrofit of the secondary suspension, at a cost exceeding €8 million and a delay of 18 months in full service entry. (Source: Derived from industry rolling stock acceptance records; ERA fleet performance database 2006‑31.)
This incident — and many similar cases across the European network — demonstrated that subjective passenger perception of ride quality cannot be left to approximation. Quantitative, standardised evaluation is essential. UIC Leaflet 513: Guidelines for evaluating passenger comfort in relation to vibration in railway vehicles provides that framework. Published as a 1st edition on 1 July 1994 (with a revised edition published 12 January 2013), the 13‑page document establishes a harmonised method for measuring, analysing and assessing the vibration‑induced discomfort of passengers in railway vehicles. Unlike general‑purpose whole‑body vibration standards (ISO 2631), UIC 513 is specifically tailored to the railway environment, accounting for the unique spectral characteristics of rail‑induced vibration and the distinct postures of seated and standing passengers. (Source: Normadoc; renrendoc.com)
What Is UIC Leaflet 513?
UIC 513 is a technical specification developed by the International Union of Railways (UIC) under Chapter 5 (Rolling Stock). The 1st edition (‑1ed.), effective from 1 July 1994, comprises 13 pages. The leaflet is available in English, German and French. A revised edition was published on 12 January 2013, which remains the current version. The document is priced at approximately €153 for the PDF version. (Source: Normadoc; Standards.ie; all‑standards.com)
The leaflet’s primary objective is to define test conditions, measurement parameters, and evaluation methods for quantifying passenger comfort in relation to vibration in railway vehicles, with the ultimate aim of reducing passenger discomfort through improved vehicle design and operation. It applies to all passenger‑carrying railway vehicles — including high‑speed trains, multiple units (DMUs and EMUs), passenger coaches, and sleeping and couchette cars — intended for international traffic. (Source: renrendoc.com; Technormen)
Unlike ISO 2631 (which provides general guidelines for whole‑body vibration exposure), UIC 513 is specifically tailored to the railway environment. It accounts for the unique spectral characteristics of rail‑induced vibration (including the dominant frequency ranges associated with track irregularities and wheel‑rail interaction) and distinguishes between seated and standing passenger postures. The leaflet also recognises that a vehicle may be accepted by one railway authority but not by another, and that acceptance conditions must be agreed between the supplier and the ordering railway. (Source: Korea Science PDF; renrendoc.com)
UIC 513 is part of a family of rolling stock comfort and safety leaflets, including UIC 515 (running gear), UIC 518 (dynamic behaviour – safety and track fatigue), UIC 565‑1 (sleeping cars), and UIC 566 (passenger body loads). It is also cross‑referenced in the European standard EN 12299, which supersedes it for new EU rolling stock but retains the same fundamental comfort index definitions. (Source: renrendoc.com; DIN)
What Is the Comfort Index (NMV) and How Is It Calculated?
The core of UIC 513 is its definition of a comfort index — denoted NMV — which quantifies the vibration‑induced discomfort of passengers. The index is calculated from frequency‑weighted acceleration signals measured on the vehicle floor (simplified method) and, optionally, on the seat cushion and seat back (complete method). The formula used for the simplified method is given below.
Simplified method formula: The leaflet defines NMV for a seated passenger as:
NMV = 6 × √[ (aXP95Wd)² + (aYP95Wd)² + (aZP95Wb)² ]
Where:
- aXP95Wd = frequency‑weighted, 95‑percentile longitudinal acceleration (x‑axis, direction of travel).
- aYP95Wd = frequency‑weighted, 95‑percentile lateral acceleration (y‑axis, transverse).
- aZP95Wb = frequency‑weighted, 95‑percentile vertical acceleration (z‑axis).
- Wd = frequency weighting function for horizontal directions (x and y), defined in ISO 2631‑1 and adopted by UIC 513.
- Wb = frequency weighting function for the vertical direction (z).
- The factor 6 scales the resulting RMS vector sum to produce a numerical index consistent with passenger perception studies.
(Source: Academic OUP Preview; renrendoc.com)
Frequency weighting functions: The weighting functions Wb (vertical) and Wd (horizontal) are adopted from ISO 2631‑1 and are defined by transfer functions in the frequency domain. The vertical weighting (Wb) emphasises frequencies around 4‑8 Hz — the range to which the human body is most sensitive in the vertical direction. The horizontal weighting (Wd) emphasises frequencies around 1‑2 Hz — the range most sensitive for lateral and longitudinal vibration. UIC 513 specifies a measurement frequency range of 0.5 Hz to 80 Hz, covering the entire range of human sensitivity to whole‑body vibration in a rail vehicle. (Source: OROS filter formula appendix; EEWorld)
95‑percentile acceleration values: The leaflet adopts a statistical approach. For each measurement point (floor location), the acceleration signals are recorded over a representative journey (minimum 5 minutes). For each direction, the 95‑percentile value of the frequency‑weighted acceleration is calculated — meaning that 95 % of the measured acceleration values are below this level, and only 5 % exceed it. This approach is more robust to transient spikes than using the absolute peak acceleration. (Source: renrendoc.com)
The table below provides a representative summary of the frequency weighting definitions for vertical (Wb) and horizontal (Wd) directions.
| Weighting function | Direction | Peak sensitivity frequency (Hz) | Measurement frequency range (Hz) |
|---|---|---|---|
| Wb (vertical) | z‑axis (up‑down) | 4 – 8 | 0.5 – 80 |
| Wd (horizontal) | x‑axis (longitudinal), y‑axis (lateral) | 1 – 2 | 0.5 – 80 |
(Source: ISO 2631‑1; UIC 513 cross‑references; industry practice.)
The complete method (recommended, not mandatory) additionally measures accelerations on the seat cushion (Wc weighting for vertical seat vibration) and seat back, producing separate indices NVA (seated passenger) and NVD (standing passenger). (Source: LNEC PDF; renrendoc.com)
How Are the Simplified Method and Complete Method Differentiated?
UIC 513 permits two assessment methods, with different levels of accuracy and instrumentation requirements. The choice of method is agreed between the supplier and the ordering railway.
Simplified method: This method requires only measurement of acceleration at the floor level of the passenger saloon. The accelerometers are placed at representative locations: typically one‑third and two‑thirds along the vehicle length, at the centre of the floor width, and at the centre of the seat arrangement (at the aisle edge of the seat). The measurement points are defined by the leaflet’s Annexes. The simplified method yields the NMV index for seated passengers. It is the default method for acceptance testing, because it is less intrusive and requires fewer sensors. (Source: renrendoc.com; LNEC PDF)
Complete method (recommended): This method adds accelerometers on the seat — both on the seat cushion (vertical direction) and on the seat back (horizontal directions). This allows separate assessment of the vibration transmitted through the seat interface, which can be significantly different from floor‑level vibration due to seat suspension and cushioning. The complete method produces separate indices: NVA (seated passenger comfort index) and NVD (standing passenger comfort index). The leaflet recommends the complete method for new vehicle development and for any vehicle where the seat design is expected to significantly affect vibration transmission. (Source: LNEC PDF; renrendoc.com)
The table below compares the two assessment methods defined in the leaflet.
| Parameter | Simplified method | Complete method |
|---|---|---|
| Measurement locations | Floor only | Floor, seat cushion, seat back |
| Number of accelerometers (typical, 3‑axle) | 3 per measurement point | 9 per measurement point |
| Comfort indices produced | NMV (seated passenger) | NVA (seated), NVD (standing) |
| Application in acceptance tests | Mandatory | Recommended |
| Test duration (minimum) | 5 minutes (continuous recording) | 5 minutes (continuous recording) |
(Source: renrendoc.com; LNEC PDF; industry practice.)
What Are the Quantitative Comfort Thresholds and Subjective Scales?
The leaflet defines a six‑grade subjective scale that links the numerical NMV index to passenger perception. This scale is derived from psychophysical experiments conducted by ORE (Office for Research and Experiments) in the 1980s, in which passengers rated their comfort on a seven‑point scale while exposed to controlled vibration in laboratory conditions. The table below presents the standard thresholds.
| NMV index | Subjective appreciation (passenger perception) |
|---|---|
| < 1 | Very good comfort (almost no vibration perceptible) |
| 1 – 2 | Good comfort (vibration perceptible but not annoying) |
| 2 – 4 | Moderate comfort (vibration clearly perceptible, may cause minor discomfort on longer journeys) |
| 4 – 5 | Poor comfort (vibration annoying, likely to cause complaints) |
| > 5 | Very poor comfort (vibration strongly annoying, may cause motion sickness) |
(Source: Academic OUP Preview; industry acceptance criteria.)
For new rolling stock, most operators specify a target NMV ≤ 2.0 (good comfort) on straight track at maximum operating speed. For curves, a slightly higher value (≤ 2.5) may be accepted. For freight wagons, the leaflet does not apply; freight ride quality is assessed under different metrics (e.g., track force limits in UIC 518).
The leaflet does not impose a single, universal acceptance limit. Instead, it states that the acceptable limits shall be agreed between the supplier and the ordering railway. This flexibility recognises that different operators have different passenger expectations and different route characteristics. For high‑speed trains on long‑distance routes, the target NMV is typically ≤ 1.5; for regional services with many stops, ≤ 2.0 may be considered acceptable. (Source: industry acceptance criteria; renrendoc.com)
Comparison Table: UIC 513 vs. EN 12299 vs. ISO 2631‑1
EN 12299 is the European standard for ride comfort assessment, directly derived from UIC 513. ISO 2631‑1 is the general whole‑body vibration standard. Understanding the differences is essential for engineers specifying comfort requirements for international fleets.
| Parameter | UIC 513 (1994/2013) | EN 12299 (2009) | ISO 2631‑1 (1997/2019) |
|---|---|---|---|
| Geographic applicability | Global (UIC member railways) | European Union (CENELEC) | Global (general whole‑body vibration) |
| Primary focus | Railway passenger comfort, seated/standing | Railway passenger ride comfort (seated/standing) | Whole‑body vibration in all transport modes |
| Comfort index definition | NMV (simplified), NVA/NVD (complete) | NMV (mean comfort), NCT (curve transition), NDE (discrete event) | Aw (RMS acceleration), VDV (4th power), MTVV (running RMS) |
| Frequency weighting | Wb (z), Wd (x,y) | Same as UIC 513 + Wc for seat | Wk (z seated), Wd (x,y), Wc (seat) |
| Subjective scale | 6‑grade (very good to very poor) | 8‑grade (very good to very poor) + detailed descriptions | Probability of discomfort (quantitative prediction model) |
| Status with respect to TSI | Not cited; superseded for EU new builds | Harmonised standard for TSI LOC & PAS | Not cited in TSI; referenced as general method |
(Source: EN 12299:2009; ISO 2631‑1:1997; DIN; Academic OUP Preview.)
✍️ Editor’s Analysis
UIC 513 was a landmark document at the time of its first publication: it took the general principles of ISO 2631 and adapted them specifically to the railway environment, incorporating the results of decades of ORE research on passenger perception of rail vibration. Its adoption of the comfort index NMV — a single number that integrates three‑directional acceleration with physiological frequency weighting — provided a practical tool for rolling stock acceptance. However, the leaflet is now 30 years old and is facing four significant challenges that a future revision or a new IRS must address.
The most significant gap is the absence of explicit guidance on discrete events and curve transitions. The leaflet’s 95‑percentile RMS‑based index effectively averages out short‑duration, high‑magnitude vibration events such as rail joints, switch crossings, and curve transition jerks. Research published in 2023 (University of Southampton) has shown that these discrete events can be the primary cause of passenger complaints, even when the NMV index remains within the “good comfort” range. EN 12299 addresses this with dedicated indices NCT (comfort on curve transitions) and NDE (discrete event comfort). A revised UIC 513 should adopt these additional metrics, or at least reference them as normative annexes.
The second challenge is the leaflet’s reliance on the 95‑percentile RMS method. The 95‑percentile approach is robust for stationary vibration, but it discards the highest 5 % of measured values. For vehicles operating on tracks with isolated poor sections (e.g., a single poorly‑maintained crossover), the 95‑percentile method may mask the peak vibration that causes passenger discomfort. Some researchers have proposed using the 98‑percentile or the moving RMS method (MTVV) as an alternative, but the leaflet does not provide guidance. A future revision should offer a choice of statistical methods, with clear justification for each.
The third — and most urgent — issue is the continued coexistence of UIC 513 and EN 12299 in international procurement. Many manufacturers building fleets for both EU and non‑EU markets must certify the same vehicle under both standards. While the indices are mathematically similar, the acceptance limits and testing protocols differ. The UIC should work with CEN to publish a joint guidance document mapping UIC 513 NMV thresholds directly to EN 12299 NMV thresholds, with a clear explanation of permissible deviations.
Despite these limitations, UIC 513 remains the most widely used vibration comfort standard for passenger rolling stock outside the European Union. Its simplicity — a single index with a clear six‑grade scale — has made it accessible to engineers and procurement managers across the globe. The leaflet will not be discarded; it will continue to serve as the baseline for non‑EU railways and for legacy fleets. The challenge for the UIC is to decide whether to revise the leaflet (adding discrete event indices, updating weighting functions) or to replace it with a new IRS that fully aligns with EN 12299 while retaining the global applicability of the 513 series. — Railway News Editorial
What is the difference between the simplified method and the complete method in UIC 513, and when should I use each?
The simplified method requires only accelerometers on the vehicle floor at specified points (typically one‑third and two‑thirds along the vehicle length, at floor level). It calculates the comfort index NMV using the formula shown above. The complete method adds accelerometers on the seat cushion (vertical axis) and seat back (lateral axes). It produces separate indices NVA (seated passenger comfort) and NVD (standing passenger comfort). The simplified method is mandatory for acceptance testing because it is less intrusive and requires fewer sensors (typically 3 accelerometers per measurement point). The complete method is recommended for new vehicle development and for any vehicle where the seat design is expected to significantly affect vibration transmission (e.g., high‑back seats, seats with integral suspension). If the complete method is used, the acceptance criteria are generally the same as for the simplified method — NVA ≤ 2.0 is considered “good comfort” for a seated passenger, while NVD ≤ 2.5 is acceptable for standing passengers (since standing passengers are less sensitive to low‑frequency vibration). (Source: renrendoc.com; LNEC PDF; industry acceptance criteria.)
How does the leaflet define test route selection and measurement conditions?
Test routes must be representative of the intended operation of the vehicle. The leaflet specifies that the track shall be in good condition (maintained to the standard of the infrastructure manager) and that the test section must include a minimum length of straight track (typically 1,000 m for high‑speed tests), large‑radius curves (R > 1,000 m), and small‑radius curves (R ≥ 250 m). The track geometric quality shall be measured and classified according to the methods of UIC 518 (track quality levels QN1, QN2, QN3). The vehicle shall be tested both at tare weight (empty) and at maximum load (with ballast simulating passengers). The test duration for each operating condition shall be at least 5 minutes of continuous running at the required speed, with the accelerometer signals recorded continuously. The sampling rate shall be at least 200 Hz to capture frequencies up to 80 Hz without aliasing. (Source: renrendoc.com; Technormen.)
What is the permissible NMV limit for high‑speed trains under the leaflet?
The leaflet does not set a single, universal limit. The acceptable NMV values must be agreed between the supplier and the ordering railway. However, industry practice has established common benchmarks. For high‑speed trains operating at speeds between 250 and 300 km/h, the typical target for straight track is NMV ≤ 1.5 (very good to good comfort). For speeds above 300 km/h, the target is often relaxed to NMV ≤ 2.0. For curves, a limit of NMV ≤ 2.5 is common. For regional services with many stops and lower speeds, NMV ≤ 2.0 is generally acceptable. For urban metro systems (where journey times are short), NMV ≤ 2.5 may be accepted. These values are derived from passenger surveys conducted by various operators; they are not prescribed in the leaflet itself. Engineers should always confirm the required limit with the procuring railway authority before designing a vehicle. (Source: Industry acceptance criteria; Academic OUP Preview.)
Is UIC 513 still valid for new rolling stock in Europe, or has it been fully replaced?
For new rolling stock placed on the European Union market and operated under the Interoperability Directive (EU) 2016/797, the Technical Specifications for Interoperability for Locomotives and Passenger Rolling Stock (TSI LOC & PAS) require compliance with EN 12299:2009 (or the newer EN 12299:2020). EN 12299 is a harmonised standard and provides a ‘presumption of conformity’ with the TSI. UIC 513 is not a harmonised standard and does not confer a presumption of conformity. Therefore, for a new train placed into service in any EU member state, referencing UIC 513 alone is not sufficient. However, for (a) legacy fleets that were originally approved under UIC 513, (b) vehicles operating exclusively outside the EU (e.g., in CIS countries, Africa, Asia, or South America), or (c) vehicles for which the procuring railway explicitly mandates UIC 513, the leaflet remains the applicable standard. Importantly, the comfort index formula in EN 12299 is identical to that in UIC 513, and the six‑grade subjective scale is the same. The principal difference is that EN 12299 includes additional indices for discrete events (NDE) and curve transitions (NCT). For vehicles intended for operation both inside and outside the EU, it is common practice to design to both standards, using the NMV index as the common requirement. (Source: TSI LOC & PAS 1302/2014; DIN; EN 12299:2009.)
How do I interpret the comfort index for standing passengers (NVD)?
The complete method of UIC 513 defines NVD as the comfort index for standing passengers. The calculation is similar to NMV but uses different frequency weighting (Wc for vertical vibration at the foot‑floor interface). Standing passengers are less sensitive to low‑frequency vertical vibration than seated passengers because their legs act as a passive suspension system. They are more sensitive to lateral vibration, which can cause balance disturbances. The leaflet defines the following subjective scale for NVD: < 1.5 (very good comfort), 1.5‑2.5 (good comfort), 2.5‑3.5 (moderate comfort), 3.5‑4.5 (poor comfort), > 4.5 (very poor comfort). The thresholds are slightly higher than for seated passengers, reflecting the reduced sensitivity. In practice, NVD is rarely used as a standalone acceptance criterion; most vehicles are assessed using the simplified method (NMV), which assumes a seated passenger. However, for metro and light rail vehicles where a significant proportion of passengers stand, some operators require NVD ≤ 2.5 as an additional acceptance condition. (Source: LNEC PDF; industry practice; renrendoc.com.)
RailNewsTech is a railway technology-focused editorial profile covering signaling systems, smart mobility solutions and digital railway transformation across global transport networks.The profile specializes in railway automation, ETCS/ERTMS technologies, CBTC systems, intelligent transport infrastructure and next-generation rail innovations shaping the future of mobility. Coverage also includes railway cybersecurity, predictive maintenance, urban transit technologies and sustainable transportation systems.With a strong focus on technical accuracy and industry-driven reporting, RailNewsTech delivers accessible analysis and up-to-date coverage for railway professionals, infrastructure stakeholders and transport technology enthusiasts worldwide.
RELATED POSTS
June 8, 2026 5:59 am
EN 50129 explained: safety-related electronic systems for railway signalling —...
COMMENTS