Foundation, Concrete and Earthquake Engineering

What is Pounding? Structural and Foundation Issues of Pounding

What is pounding?

Collision of adjacent buildings during earthquake is generally called pounding. This is occurred when they have different dynamic properties and there have not sufficient or no separation distance from adjacent one. The damage due to pounding may be of local and global; local damage is associated with collision force but global damage of buildings depends on their dynamic properties at the time of taking place of collision. Global damages caused by transfer of energy and momentum between two system generated by collision.

Seismic pounding between buildings is one of the most common causes of severe structural damages due to earthquake. Pounding involves movement along or transverse to separation joints provided between neighbor structures and may cause non-structural damages. From the above discussion, seismic pounding between adjacent buildings occur during earthquake when

• Buildings have different dynamic characteristics

• They vibrates out of phase relative to each other

• The separation between buildings at rest is not sufficient

• Insufficient energy dissipation system, if any, to restrain movement within allowable separation distance. 

Why is pounding of buildings concerned?

Wall collapse due to pounding during Loma Prieta earthquake (1989)
Figure 1: Wall collapse due to pounding
during Loma Prieta earthquake (1989)

Past building codes didn’t provide definite recommendation or guidelines to account pounding effect and to counteract this phenomenon. As a result in many densely populated areas to achieve maximum land usage and for economic consideration, many buildings over the world already built extremely close to neighbor and in some region even no space is left. These building are vulnerable to pounding damage during future seismic activity. A large separation distance is not expected from both technical point of view within same facility having large expansion joints and economical point of view considering loss of land. In many cities the highly congested structures become a major concern for pounding damage. This is why, it is now widely accepted that this unexpected phenomenon has to be mitigated or prevented (Abdel Raheem, 2006).

What is seismic gap?

The distance provided between two adjacent building structures is called separation joint; often same facility is divided into two wings and sometime more depending on dimension of buildings. This permits an independent movement of structures relative to each other. A seismic gap is nothing but a separation joint kept to provide room for relative lateral movement due to seismic agitation.

Considering functional continuity, building utilities have to be extended from one wing to other across the building separation and finishing is provided for architectural termination on either portion. This separation joint for older buildings may be only one or two inches. For newer buildings these separation may be as much as one foot based on desired horizontal movement or seismic drift. All details about flashing, piping, HVAC ducts, fire sprinkler facility, flooring and partitions have to be done to permit two wings to move to expected distance at these locations when two structure or separated wings more closer to each other or move apart during earthquake. Damage of such non-structural elements across seismic gaps is very common. When these gap is inadequate, pounding between buildings may cause damage to structural elements of these buildings.


Many events of seismic pounding have been recorded to date. Pounding event has resulted worse damage and many cases of collapse of entire building. The earthquake that rocked Mexico City (1985) has disclosed (Rosenblueth and Meli, 1986) that pounding damage was present in more than 40% of total severely damaged or collapsed building out of 330 buildings surveyed and 15% of that resulted collapse. This earthquake left behind lesson about significance of pounding by leaving maximum number of building damage till that date (Bertero, 1986).

Not only that, past and recent earthquake brought our several instances of such damages to both bridge and building structures. Some significant earthquakes are saguenay earthquake (Canada) in 1988, Cairo earthquake (Egypt) in 1992, Northridge earthquake (USA) in 1994, Kobe earthquake (japan) in 1995. Kocaeli earthquake (Izmit, Turkey) and Tohoku earthquake (japan) in 2011. Some example of these earthquake are listed below:

Pounding damage in the 1985 Mexico City earthquake
Pounding damage in the 1985 
Mexico City earthquake

Cracking due to Pounding (Loma Prieta Earthquake, 1989)
Cracking due to Pounding
(Loma Prieta Earthquake, 1989)

Pounding between wooden and reinforced framed structure in Cairo, 1992
Pounding between wooden and reinforced 
framed structure in Cairo, 1992

Pounding damage of Northridge earthquake (USA, 1994)
Figure 2: Pounding damage of Northridge
 earthquake (USA, 1994)

Rigid architectural flashing results pound of two wings of buildings (Northridge earthquake, 1994)
Rigid architectural flashing results 
pound of two wings of buildings 
(Northridge earthquake, 1994)

Pounding damage in 2015 Nepal earthquake
Pounding damage in 2015 Nepal earthquake

Classification of pounding damage pattern

Pounding damage of building was investigated after Loma Prieta earthquake (1985) in San Francisco bay region; based on this damage patterns of pounding were classified into four types (Kasai and Maison 1991), they are

• Type-1, where major damage of buildings were reported Fig 1

• Type-2, a life threatening hazard was created by falling or failure of building accessories

• TYPE-3, building functionality is lost as important electrical,mechanical or fire protection systems become out of order.

• TYPE-4, this involves architectural and/or trivial structural damage; examples of each type will be discussed in pounding survey section.

What are the pounding vulnerable structures?

• Comprehensive pounding damages were reported in ‘unreinforced low-rise buildings’ having no building separation

• Modern buildings are well designed for seismic safety, but often may subjected to wrong architectural details which generally includes building separation is filled with rigid architectural flashings (Cole and Takewaki, 2011).

• Condition becomes worse when adjacent buildings have dissimilar dynamic characteristics and vibrate out of phase during earthquake and sufficient separation is not provided.

• Or insufficient energy dissipation system to restrict building to move within designed gap between adjacent facilities.

In the past earthquake, it was noticed that significant damage was encountered within the perimeter of 90 km around epicenter which is an indication of disastrous damage under earthquake event at sites close to epicenters. A details description of vulnerable structures are presented below:

Why are URM buildings most vulnerable to pounding?

Special attention to unreinforced masonry (URM) buildings is required as pounding damage of such buildings found sufficiently frequently and a common damage pattern is outlined from experiences in Christchurch earthquake (2011). The cracking pattern was found The cracking pattern was found to extend through masonry walls typically from topmost point of contact of two building to lintel or window arch whatever found nearer in either building. Then cracks often extend from window opening and reached up to top of parapets of buildings. But parapets damage were not generally linked to effect of pounding. URM building not having ductile steel skeleton is susceptible to pounding damage.

What is gel/space ratio? Its significance in concrete strength prediction

Gel/space ratio

The ratio of volume occupied by hydrated cement paste to the aggregated volume of capillary pores and hydrated cement paste is known as gel/space ratio. It is denoted by r. Powers (1958) found that compressive strength of concrete is 34000 r3 psi (234 r3 MPa) and interestingly he found no influence of mix proportion of concrete and age of it on strength prediction. To realize the definition and significance of gel/space ratio it is required to discuss about volume of hydration product.

Volume of hydration produts

The total space available to occupy by productsof hydration is the summation of absolute volume of fresh cement and the volume of mixing water. Of these, if small loss of water under the contraction of the cement paste and that due to bleeding is ignored, the water consumed by chemical reaction with C2S and C3S was found to be 21 and 24 percent (very roughly) of the mass of two respective silicates. If the final reaction of hydrate C4AF is
The respective figures of C3AF and C3A are 37 and 40 percent. Equation (1) is also vary approximately due to our inadequate knowledge of stoichiometry of the hydration products and cannot be ascertained the amount of chemically combined water. 
Gel formation in cement paste (Paul Bennison)
a) Cement paste just after mixing
b) Initial stiffening-starting reaction after mixing
c) Initial hardening-formation of physical structure
d) Later gradual hardening-infilling pores with gel
Non-evaporable water determined under specific conditions is considered as 23% if anhydrous cement (measured by mass); in case of type II, moderate sulfate resistant cement, this value may be 18%. The specific gravity of hydration products of cement becomes such that the resulting volume is more than absolute volume of anhydrous cement.
Volume of hydration product of cement
The average value of specific gravity of hydration product in saturated structure, inclusive of pores available in the possible densest structure, is 2.16. Here we are providing a demonstration of calculation of volume change during hydration.

Example 1.0

Mass of cement=100 g
Specific gravity of cement=3.15
That is absolute volume of hydrated cement=100/3.15=31.8 ml
Volume of non-evaporable water= 23ml (23% of mass of cement)
Volume of solid product of hydration of cement=31.8+0.23 X 100 (1-0.254)
=48.9 ml

The cement paste at this condition has characteristic porosity around=28%

That is 

Where wg=volume of gel water 
i.e, wg=48.9X0.28+0.28Xwg
or, wg=19ml

Thus the volume of hydrated cement paste=48.9+19=67.9 ml 

The summary of the example 1 can be drawn as in table-1
Volume of hydration product
From table -1 it can be concluded that
Total volume of water in the mix=23+19=42 ml
Water/cement ration= 42/100=0.42 (by mass)

Water/cement ration= 42/31.8=1.32 (by volume)
Actual volume cement+water=31.8+42=73.8 ml

Volume reduction during hydration=73.8-67.9= 5.9 ml
Volume of hydration products for 1 ml of dry cement=67.9/31.8=2.13 ml

Example 1.0 is ideal condition, where hydration is assumed to occur in sealed container where no water movement is allowed whether into or away from the system. It is interesting to note that reduction in volume by value 5.9 stand for empty capillary pores dispersed in the hydrated cement paste.

From the example 1.0, it is found that after hydration of cement, it occupies two times of its actual volume (i.e. 2.13 ml of 1 ml cement). Because all products of hydration are not gel, in the following discussion we assume hydration of 1 unit cement will occupy about 2.06 units (volumetric unit). From the definition of gel/space ratio 
definition of gel/space ratio

C = Mass of cement in mix

Vc = Specific volume or volume occupied by unit mass of cement

f = Fraction of cement hydrated
wo=volume of water in the mix 

  If specific volume of cement in dry state is considered 0.319 ml/gm above equation becomes 
Relation between strength and water/cement ratio we know a reasonably low water/cement ratio will produce more strength, Too low water cement ratio may reduce workability and may supply insufficient water to hydrate significant fraction of cement incorporated in the mix. A too high water/cement ratio produces a porous structure leaving concrete of inferior quality both strength and durability point of view. But this relationship with water/cement ratio does not based on a well constituted law as water/cement ratio rule cannot be validated by proper qualification. Strength of concrete at any water/cement ratio depends on other factors like

  • Physical and chemical properties of cement
  • Degree of hydration of it
  • Ambient environment around hydration process like air content in the concrete and temperature. 
  • Change in effective w/c ratio
  • Crack formation due to bleeding 
  • The cement content in the mix
  • Properties of interface between aggregate cement paste


Relation between strength and gel-space ratio 

As water/cement ratio is not sole factor to control strength of concrete, it is more accurate to relate strength with the distribution and concentration of solid materials produced from hydration of cement in the available space for these materials. Power (1946) has first determined the relation between development of strength and gel/space ratio of mortar has shown in following figure.

From the figure, it is noticed that strength has approximate proportionality with cube of gel/space ratio and he figured out that 234 Mpa was the inherent strength of gel for that type of cement and specimen examined at that time. For usual Portland cement we used in our construction works, the numerical vales were found more or less same; some exceptions are observed for cements having higher C3A content as they yield lower strength at a particular gel/space ratio.

Correction for specific gravity of adsorbed water

In discussing this topic, some question need to be answered, these are:

Why specific gravity of adsorbed water is 1.1?

What is disjoining pressure?

An investigations applying nuclear magnetic resonance showed that gel water required same energy to make bond as that remain in interlayer water of some expansive clays; 

Effect of temperature on gel/space ratio

We know a higher temperature at the time of placing of concrete and maintaining or allowing during setting, definitely increase strength at very early stage; followed by often adverse effect on strength at and after 7 days. This is due to rapid initial reaction that occur during hydration leading to formation of hydration product of poorer quality; a poor physical structure having more pores.Thus a significant portion of pores will always leave unfilled. According to gel/space ratio rule, the gel/space ratio will be lower in this case as compared to concrete that is hydrated slowly. The latter one will have less pores, thus resulting higher gel/space ratio.

According to Verbeck and Helmuth, the explanation of low gel/space ratio due to high early temperature and subsequent adverse effects on later strength of concrete, is higher rate of hydration at initial stage under higher temperature inhibit successive hydration. . This is due to an irregular distribution of hydration products in the paste. Diffusion away of the hydration product, in this case, takes inadequate time from cement particles under initial higher temperature and cannot produce uniform precipitation into interstices of paste.

Thus high density of hydration product is produced in vicinity of particles still waiting for hydration reaction or in state of some degree of hydration and inhibit subsequent hydration. As a result gel/space ratio will be less in the interstices and adversely y affects strength of concrete in the long run. Thus the local weakness in the concrete section will permit to progress cracks and results a lower strength of it as a whole. With the help backscattered electron imaging it has been confirmed that porous C-S-H exist in between hydrated cement particles.

Influence of chemical composition of cement

As per chemical composition, only gypsum content of cement is here prime concern. For a given cement, the gypsum content defines shrinkage and setting time of cement; but the gypsum content that is satisfactory in respect of shrinkage is found not adequate for establishing desired setting time. Gypsum is added to cement clinker during manufacturing process of cement and then final grinding of clinker is done. As discussed in above section about temperature, the initial structure produced during setting determines structure of cement paste at upcoming stages of hydration. Thus gel/space ratio is dependent on optimum gypsum content to facilitate hydration of as much as cement particles to gain strength and behavior of concrete under shrinkage and creep are also affected.