# Vibration damage to structures

> One page of the phonometry documentation, as Markdown. Index: https://docs.jmrp.io/phonometry/llms.txt

Canonical: https://docs.jmrp.io/phonometry/vibration/structural/structural-damage/
Language: en
Alternate: https://docs.jmrp.io/phonometry/es/vibration/structural/structural-damage/index.md
Version: phonometry 4.0.0
Install: pip install "phonometry[full]"
Author: José Manuel Requena Plens (https://orcid.org/0000-0003-1250-6212)
Published: 2026-09-08
Updated: 2026-11-16T21:36:14Z
Standards: DIN 4150-3:1999-02
Verified: 27 conformance checks (DIN 4150-3:1999-02) https://docs.jmrp.io/phonometry/reference/conformance/#vibration-effects-on-structures-din-4150-3
License: MIT (https://github.com/jmrplens/phonometry/blob/main/LICENSE)
Summary: DIN 4150-3 guideline values: peak particle velocity by building class and frequency, buried pipes, long-term values and the bending stress.

---

Everything else in this section follows vibration because it ends up as
**sound**. This page follows it because it ends up as a **crack**. A pile
driver next door, a tram at the end of the street or a blast on a nearby site
puts vibration into the ground, the ground puts it into a foundation, and the
question the neighbours ask has nothing to do with audibility: will the
building be damaged?

Answering it properly means the dynamic stresses in the structure, compared
against what the material and the design allow. DIN 4150-3 offers the cheap
answer first, and it is the one used in practice: **guideline values**
(*Anhaltswerte*) for a single measured quantity, the peak particle velocity,
drawn from a large body of measurements on real buildings. Keep under them and
damage of the kind the standard defines has not been observed in that body of
measurements, which is a statement about the evidence rather than a promise
about this building: meeting a guideline value satisfies the DIN 4150-3
criterion and certifies nothing. Exceed them and nothing follows automatically. The standard is explicit that damage does not
have to occur; what has run out is the cheap answer, and the stress
calculation has to be done instead.

## 1. What is measured

Clause 5.1 fixes the quantity before it fixes any number. At the **foundation**,
the three components $v_x$, $v_y$ and $v_z$ of the particle velocity are
recorded separately and the largest peak of the three, $|v_i|_\mathrm{max}$, is
the value judged; the standard then writes it $v_i$ and so does this page. The
transducers go in the lowest storey, on the foundation or the outside wall, and
on an unbasemented building no more than 0.5 m above ground level, preferably
on the side facing the excitation.

In the **topmost floor plane** the quantity changes: the larger of the two
horizontal components, measured in or close to the outside wall. That
measurement is not a second excitation. It is the building's horizontal answer
to the one at its foundation, which is why the guideline value there does not
depend on frequency at all: the response is already at the structure's own
frequencies.

![Section through a three-storey dwelling with no basement, with a piling rig on the ground to its left and dashed waves running from the pile through the ground toward the building. Four numbered points are marked. Point 1 is a transducer in the lowest storey, on the inner face of the outside wall that faces the source, just above the ground floor slab. Point 2 is on the same outside wall just under the top slab, with arrows for the two horizontal directions x and y, x pointing out of the building toward the source. Point 3 stands on a floor at mid-span, the span split into two equal halves, with an arrow for the vertical direction z. Point 4 is a buried pipeline drawn in cross-section running past the building, more than 2 metres from the outside wall and dimensioned as such, bared at one spot with a transducer on its crown. Below, two enlarged details. The first shows the foundation point with its three directions close together on the inner face of the outside wall, x pointing out toward the source, under a dashed line 0.5 m above the ground level outside. The second shows the pipe along its axis with the transducers on the pipe itself, z, y and x recorded and x along the axis, a dashed substitute point on the ground above it marked as only an estimate, usually larger than on the pipe, and the pipe judged on Table 2 by material. Three boxes at the foot say what is read at each building point: the largest peak of the three components at the foundation, against Table 1 at the frequency of that peak; the larger horizontal peak in the topmost floor plane, against one value per class; and a vertical velocity of 20 millimetres per second or less on the floor, for short-term vibration. The lines underneath give a reading of 5.1 millimetres per second at 16.5 hertz at the foundation of a dwelling against the 6.6 read off Bild 1, the meter and its calibration, the frequency and acceleration limits for a transducer set down loose, and when several points or storeys are measured at once](https://docs.jmrp.io/phonometry/media/diagram_structural_damage_points.svg)

### How the measurement goes

Use a vibration meter that meets DIN 45669-1, function-checked and calibrated at
a reference frequency and amplitude for the job. At the foundation, three
transducers go close together in the lowest storey, on the foundation of the
outside wall, on the wall or in a recess of it, preferably on the side facing the
source and, without a basement, no higher than 0.5 m above ground level. They
record $z$ and two horizontal directions at right angles, one parallel to a side
wall and preferably $x$ pointing at the source; a large footprint takes several
points at once. Loose steps, loose slabs and hollow-sounding surfaces are
unsuitable, and on a hard surface a transducer set down loose holds only to
100 Hz vertically and 40 Hz horizontally at 3 m/s² or less; otherwise it is
glued, screwed or plastered on. In the topmost floor plane, $x$ and $y$ go in or
close to the outside wall; where relevant, on the floors expected to vibrate
most, $z$ goes at about mid-span. On a pipeline the transducers go on the pipe
where possible, bared only at the measuring point, one horizontal direction along
the axis; the ground above is a substitute where the source is neither close
beside the pipe nor a short distance below it, and gives only an estimate,
usually larger. Record the whole event, or representative stretches, and where
possible judge extraneous vibration by comparing the readings with and without
the source acting. Read the peak of each component and its frequency, and report
what Annex A lists: every point, its direction and its distance to the source.

## 2. Short-term vibration, and why the guideline rises with frequency

Short-term vibration is vibration that does not occur often enough to build up
resonance in the structure: construction work, blasting, a passing train on a
line that is not busy. Table 1 gives its guideline values for three building
classes, and at the foundation they depend on frequency.

They rise, and the reason is that the standard is really about strain. The
same velocity at a lower frequency means a larger displacement, and it is the
displacement across a wall that opens a crack in the plaster. So a building
tolerates a fast wiggle better than a slow one, and the guideline value climbs
from 1 Hz to 100 Hz. Above 100 Hz the 100 Hz value may be used.

![Three panels. The large left panel plots the guideline peak velocity against frequency from 1 to 100 hertz for the three building classes: each curve is flat to 10 hertz and then rises in two straight segments, commercial and industrial from 20 to 50 millimetres per second, dwellings from 5 to 20, and especially sensitive buildings from 3 to 10, with the area below the lowest curve shaded. The top right panel gives the topmost floor plane values as paired horizontal bars, short-term against long-term: 40 against 10, 15 against 5, and 8 against 2.5 millimetres per second. The bottom right panel gives the buried pipeline values the same way: welded steel 100 against 50, concrete and flanged metal 80 against 40, masonry and plastic 50 against 25.](https://docs.jmrp.io/phonometry/media/structural_damage_guidelines.svg)

**Show the code for this figure**

```python
import matplotlib.pyplot as plt
import numpy as np
from phonometry import vibration

freq = np.linspace(1.0, 100.0, 400)

fig, ax = plt.subplots()
for name in vibration.BUILDING_CLASSES:
    ax.plot(freq, vibration.guideline_velocity(name, freq), label=name)
ax.set(xlabel="Frequency [Hz]", ylabel="Peak velocity $v_i$ [mm/s]")
ax.grid(True, alpha=0.3)
ax.legend()
plt.show()

# One measurement against the guideline it is judged by, drawn on the same
# curves:
res = vibration.assess_building_vibration(
    4.2, building_class="residential", frequency_hz=18.0
)
print(round(res.guideline_mm_s, 2), res.within_guideline)   # 7.0 True
res.plot()
plt.show()
```

*The three tables at once. The left panel is Bild 1: the guideline value at
the foundation, flat to 10 Hz and rising after it. The right panels are the
values that carry no frequency, each with the long-term value beside it.*

Table 1 prints the middle band as a range, "5 bis 15" for a dwelling between
10 Hz and 50 Hz, and a range does not say what 30 Hz is worth. Bild 1 says it:
the corner values joined by straight lines on a **linear** frequency axis. A
dwelling at 30 Hz is allowed 10 mm/s, and that is a reading of the figure, not
of the table.

```python
from phonometry import vibration

# The corners Table 1 prints, for a dwelling.
for f in (1.0, 10.0, 50.0, 100.0):
    print(f, vibration.guideline_velocity("residential", f))
# 1.0 5.0 / 10.0 5.0 / 50.0 15.0 / 100.0 20.0

# Inside a band, which only Bild 1 decides.
print(vibration.guideline_velocity("residential", 30.0))          # 10.0

# In the topmost floor plane there is nothing to read against frequency,
# and passing one anyway is refused rather than ignored.
print(vibration.guideline_velocity("residential", location="top_floor"))  # 15.0
```

The three classes are the rows of Table 1: `"commercial"` for commercial and
industrial buildings and buildings of like structure, `"residential"` for
dwellings and buildings of like construction or use, and `"sensitive"` for
buildings that fit neither and are worth preserving, a listed building being
the example the standard gives. A massive engineering structure such as a
reinforced-concrete abutment or a block foundation may raise the row 1 values
by up to a factor of two, which is what 5.1 allows provided nothing dangerous
comes of the soil mechanics. The allowance is a sentence of Clause 5 naming
Table 1, so it stops at short-term vibration and the library refuses to carry
it over to Table 3:

```python
print(vibration.guideline_velocity("commercial", 50.0, massive_structure=True))
# 80.0
```

Every choice on this page is a `Literal` in the signature: the building class,
the place of measurement, the duration and the pipe material. A type checker
therefore flags `"residental"` before the call runs, and the call still refuses
it at run time with the three names it accepts.

## 3. Floors, and the one number that covers them

A floor or ceiling gets its own sentence rather than a table. Clause 5.2: if
the vertical velocity at the point of largest vibration, in general mid-span,
stays at or below 20 mm/s, no reduction in the serviceability of the floor is
expected. The value is published as
[`FLOOR_VERTICAL_MM_S`](https://docs.jmrp.io/phonometry/reference/api/vibration/building-damage/),
and for a building in row 3 of Table 1 the standard notes that even that may
have to be reduced.

## 4. Buried pipelines

A pipe in the ground is judged on the pipe, not on the building above it, and
by what it is made of. Table 2 has three rows, and Clause 6.3 lets long-term
vibration use the same table at 50 % without further evidence:

```python
from phonometry import vibration

print(vibration.pipeline_guideline_velocity("welded_steel"))                # 100.0
print(vibration.pipeline_guideline_velocity("masonry_or_plastic"))          #  50.0
print(vibration.pipeline_guideline_velocity(
    "masonry_or_plastic", duration="long_term"))                            #  25.0
```

House connections up to 2 m from the building are judged by the building's own
foundation values instead, and drainage pipes by row 3 whatever they are made
of.

## 5. Long-term vibration

Long-term vibration (*Dauererschütterungen*) is the opposite case: it occurs
often enough for the structure to respond at its own frequencies. Table 3
answers it with one value per class in the topmost floor plane and nothing at
the foundation, because a structure ringing at its own frequency is judged
where it rings. The values run from a quarter to a third of the short-term ones, and the
fraction is not the same for the three classes: 10 against 40, 5 against 15 and
2.5 against 8 mm/s.

```python
from phonometry import vibration

for name in vibration.BUILDING_CLASSES:
    short = vibration.guideline_velocity(name, location="top_floor")
    long = vibration.guideline_velocity(
        name, location="top_floor", duration="long_term"
    )
    print(name, short, long)
# commercial 40.0 10.0 / residential 15.0 5.0 / sensitive 8.0 2.5
```

Asking for a long-term value at the foundation raises, because Table 3 does not
have that column.

## 6. From a velocity to a stress

Clause 6.2 is the bridge back to the proper calculation. For a beam or a
one-way slab of full rectangular section, constant stiffness and uniform mass,
vibrating in one mode, the peak bending stress follows from the peak velocity
alone:

$$
\hat{\sigma}_\mathrm{max} = 1.73 \left(E_\mathrm{dyn}\, \varrho\,
\frac{G_\mathrm{ges}}{G_\mathrm{balken}}\right)^{0.5} k_n\, \hat{v}_\mathrm{max}
$$

No length, no depth, no span: the dimensions cancel, which is the whole reason
the formula is worth having. A velocity measured where the amplitude is largest
is enough. The mode coefficient $k_n$ lies between 1 and 1.3 in the
technically important cases, so it can move the answer by less than a third,
and $G_\mathrm{ges}/G_\mathrm{balken}$ is 1 for a beam carrying nothing but
itself.

```python
from phonometry import vibration

# Concrete, 30 GPa dynamic modulus, 2400 kg/m3, first mode, 10 mm/s peak.
# Note the unit: the tables are in mm/s and this formula is in m/s.
sigma = vibration.bending_stress(
    0.010, dynamic_modulus_pa=3.0e10, density=2400.0
)
print(round(sigma / 1e6, 2), "MPa")     # 0.15 MPa
```

Clause 6.4 adds the estimate that says whether the topmost floor plane will
answer at all: for a building of about five storeys or more, the lowest
horizontal natural frequency is roughly $f_i \approx 10/n$ with $n$ the number
of storeys.

```python
print(vibration.story_fundamental_frequency(10))    # 1.0 Hz
```

Below that, `BuildingDamageWarning` says so: four storeys still return
2.5 Hz, because "about five" is not a line the standard drew, but the number
is an extrapolation of a rule offered for taller buildings rather than the
rule itself.

## What this guide covers

- **Covered.** The **guideline values** of DIN 4150-3 as published: Table 1 at the foundation
  against frequency and in the topmost floor plane, Table 2 for buried
  pipelines, Table 3 for long-term vibration, the doubling of row 1 for massive
  engineering structures (5.1), the 50 % reduction of Table 2 for long-term
  vibration (6.3) and the 20 mm/s vertical value for floors (5.2).

- **Covered.** The reading of **Bild 1** between the printed corner values: straight lines on
  a linear frequency axis, so a value inside a band is defined rather than left
  to the reader.

- **Covered.** The **bending stress** of Formula (1) in 6.2 and the storey estimate
  $f_i \approx 10/n$ of 6.4.

- **Not covered.** **Nothing predicts the vibration**: the guideline values are compared against a
  measurement you supply. Prediction of ground-borne vibration from a source to
  a building, which is DIN 4150-1, is not here but in [Predicting vibration
  before measuring](https://docs.jmrp.io/phonometry/vibration/immission/vibration-prediction/), and
  neither is the human perception of vibration in buildings of DIN 4150-2, which
  [Vibration and people in buildings](https://docs.jmrp.io/phonometry/vibration/immission/people-in-buildings/)
  assesses; the meter of DIN 45669-1 is in [Measuring vibration
  immission](https://docs.jmrp.io/phonometry/vibration/immission/vibration-meter/). For people
  rather than structures, see also the [human
  vibration](https://docs.jmrp.io/phonometry/vibration/human/) pages.

- **Not covered.** **The stress route is one formula, not a design check.** Clauses 4.2 to 4.4,
  which determine the dynamic stresses by measurement or calculation and compare
  them against the permissible ones, are what a reader has to do when a
  guideline value is exceeded; only the closed-form bridge of 6.2 is implemented.
  Fatigue, settlement of non-cohesive soils under repeated vibration (4.6, Annex
  C) and foundation displacement are described in the standard and not modelled
  here.

- **Not covered.** **The measurement is described, not checked.** Where the transducers go, how
  they are set down and what the report carries are 5.4, Annex A and the
  DIN 45669 parts 5.4 refers to, set out above for the person measuring. None of
  that is implemented here, and neither is the instrument itself.

## See also

- [Evaluating machine vibration (ISO 20816-1)](https://docs.jmrp.io/phonometry/vibration/machinery/machine-vibration-evaluation/):
  the same shape of question asked of a machine rather than of a building, and
  answered by evaluation zones instead of guideline values.
- [Human vibration exposure (ISO 2631, ISO 5349)](https://docs.jmrp.io/phonometry/vibration/human/human-vibration/):
  what the same vibration does to a person, weighted for the body rather than
  for the structure.
- [Transfer stiffness of resilient elements (ISO 10846)](https://docs.jmrp.io/phonometry/vibration/structural/transfer-stiffness/):
  the elements inserted to keep the vibration out in the first place.

## References

- **Deutsches Institut für Normung. (1999).** *Erschütterungen im Bauwesen — Teil 3: Einwirkungen auf bauliche Anlagen* (DIN 4150-3:1999-02).
  The measured quantity of 5.1, the short-term guideline values of Table 1 with the curves of Bild 1, the ceiling rule of 5.2, the buried-pipeline values of Table 2, the long-term values of Table 3 and the 50 % reduction of 6.3, the bending stress of Formula (1) in 6.2 and the storey estimate of 6.4. The measuring points, directions and report of 5.4, the pipeline notes of D.1 and the worked dwelling of D.2 are prose, set out in the measurement section.
- **Deutsches Institut für Normung. (2005).** *Messung von Schwingungsimmissionen — Teil 2: Messverfahren* (DIN 45669-2:2005-06).
  Where the transducers of a building measurement go and how they are set down: the foundation positions of 5.1.2, the measuring directions of 5.2, the loose-mounting limits of 5.3.2, the measurement time of 6.1, the check before use of 6.3 and the comparison with the source stopped of 7.2.2.
