Here is one of the best explanations, I've found,  on how the body becomes injured following an auto accident.  This is taken from Spine Research Institute of San Diego Module 1 Whiplash Advanced Topics.

Sequence of Events in the Ideal LOSRIC Model

LOSRIC (low speed rear impact crash/collision)

....so far we've looked at improved mathematical models (MADYMO), animal studies, state of the art ATDs (BioRID I), cadaver specimen crash tests (PMHS), and, of course, full scale human volunteer crash tests employing sleds in some cases, and real vehicles in others. I have reviewed nearly all of the current literature, omitting only a few which were so plagued with serious methodological flaws that it served no purpose to dissect them.

From all of this work we can effectively reconstruct an accurate sequence of events in a typical LOSRIC, without the cumbersome and tedious citing of specific references. The following is thus a composite of all of the relevant literature, relying heavily on the work of Ono et al. (472) and Siegmund et al. (497). Bear in mind, of course, that special risk variables can significantly alter the kinematic response to CAD trauma. And this model is based on an ideally positioned occupant, in a relatively good state of health, looking straight forward, wearing restraint belts, using a head restraint, and seated in a standard car seat and in a car struck squarely from the rear (180°) with good bumper alignment (i.e., no over-ride or under-ride, and no offset), free runout 
(i.e., no second collisions). The model is also applicable to speed changes of about 2.5 mph to 6 mph. I'll refer to this model as the Ideal LOSRIC Model. Later, we'll discuss specific known risk factors which can be used to refine our understanding of the kinematic responses of specific (i.e., atypical) crashes.
 

Phase 1 (0-50 msec)

After impact, the car seat pushes forward against the occupant. Most seat backs will yield elastically from a typical starting inclination of 20-25° to about 30-35°. The head's inertia holds it initially in space. At about 25 msec after impact, the head and T1 begin to accelerate, the head slightly faster than T1. The thoracic and cervical spines begin to straighten, resulting in an upward motion of the torso, which is enhanced by the ramping up of the torso against the seat back. This results in high compressive forces in the neck. Head rotation begins about 50-70 msec after impact.

Phase 2 (50-100 msec)

During this phase the neck assumes the s-shaped configuration, with hyperextension occurring in the lower cervical spine (chiefly C5-6), while flexion occurs in the upper spine. Due to compression of the spine, spinal ligaments become slack, thereby reducing the stiffness of the spine up to 70%. Maximal compression occurs 
around 50 msec, followed quickly by an increasing shear. The s-shaped curve is associated with the head lag period and the spine is subjected to high shear force. All of this happens prior to the head strike against the head restraint at about 90-120 msec--if it is in good enough position. This is when peak head horizontal acceleration (about 2.5 x the acceleration of the vehicle) and peak head over-speed (about 2 times the speed change of the vehicle) occurs. By 80-110 msec the head is beginning to move relative to the earth, but by this time up to two inches of retraction has occurred between the head center of mass and T1. Most of the injury probably occurs during this phase.

Phase 3 (100-150 msec)

Torso extension reaches a maximum around 130 msec and extension of the head begins. Ramping motion reaches its maximum at about 150 msec, coinciding with cervical spine extension. Depending on the occupant's starting position, relative position of the head restraint, interaction with the seat back, and stature, injury mechanisms can continue through this phase.
 

Phase 4 (150-300 msec)

The head/neck extension angle become maximum at about 250 msec, but this is largely dependent upon the occupant's starting position, relative position of the head restraint, interaction with the seat back, stature, and the acceleration level. Accordingly, some potential for injury exists in this phase as well. The general 
sequence of events is portrayed in Figure 1.4. Here the s-shaped curve configuration is illustrated, with hyperextension of the lower segments and flexion of the upper segments. This is probably the chief point of injury and occurs before the head strikes the head restraint.
 

Neck Pain

Tearing or damage to any soft tissue (including nervous tissue), fracture of bone, or disc herniation/prolapse/protrusion/disruption can cause neck pain. Immediate pain often indicates more severe injury, but many disabling injuries have delayed onset of symptoms. Very early onset of severe pain is sometimes an indication of disc or ligament injury. The balance of current evidence implicates paraspinal soft 
and hard tissue as the chief locus of pain generation. This includes discs, ligaments, joint capsules, end-plates, neural tissue, and the vertebrae themselves. Moreover, although the muscles are often found to be in spasm, sore, and tender, and treatment rendered to the muscles does provide some symptomatic relief, they are not usually the source of pain. This is a departure from popular thought, and represents also a 
change in my teaching. However, Mooney (539) provides an argument in favor of muscles as an injured or directly involved tissue. He notes that cervical muscles contain the highest concentration of muscle spindles--up to 500 per gram of muscle--and these contain intrafusal fibers which enhance sensitivity. These have afferent and efferent connectivity to the sympathetic nervous system.

Another interesting new study (540) found that patients with late whiplash have a sense of pain, stiffness, and tension in their muscles as a result of an inability to relax between repetitive contractions. A similar finding has been reported in fibromyalgia patients and in those with work-related trapezius myalgia. In addition to 
somatic complaints of neck, back, and shoulder pain, a high proportion of the patient group had cognitive dysfunction. A variety of explanations for the muscular abnormalities was proffered: 1. the vicious cycle of pain, 2. sensory activity from ligaments and joint capsules increasing muscle spindle activity, and 3. the 
increased muscle metabolites stimulating group 3 and 4 muscle afferents, which activate both static and dynamic gamma motoneurons projecting to homonymous as well as heteronymous muscles. And, not to leave out the psychological-stress-muscle tension theory!

Another recent paper (541) reported a higher tenderness and sensitivity to palpation and algometer in whiplash patients compared to controls. The authors conjectured a disturbance of pain modulation. They stated that a, "central dysmodulation of nociceptive impulses does not necessarily imply a central lesion caused by the accident. It might be induced by a long-lasting peripheral noxious input with sensitization of central synapses."

Barnsley et al. (169t) have made a convincing case for the prominence of facet joint pain as the culprit in a large percentage of chronic whiplash pain patients by injecting local anesthetics into the joints. Painful joints were identified in 54% of their patients. Having identified the origin of the pain in those patients, it would have 
been interesting to compare a trial of manipulation on them to determine how much of that pain was biomechanical. More recently, using the same technique, these authors reported that between 46% and 73% of their CAD patients had pain localized to the cervical zygapophysial joint or supporting ligaments. In 26 of 31 patients, a single segment was the causative factor, with C2-3 the most common cause of headache and 
upper cervical spine pain, and C5-6 the most common level associated with lower cervical spine pain and referred arm pain--a particularly interesting finding in light of the recent work by Ono et al. (472), Grauer et al. (496), and Panjabi et al. (504), who measured peak hyperextensions in lower cervical levels, while upper 
levels rotated into flexion. Jónsson et al. (169p) reported the highest proportion of ligamentous instability at this level in their series of CAD patients. I've made a composite illustration of the types of lesions reported to occur in CAD trauma. These are illustrated in Figure 1.6. In this illustration, a sagittally semi-sectioned slice of 
cervical spine (black portion of inset) is labeled with the various reported lesions. Used with permission from 
Whiplash: A Patients Guide to Recovery, San Diego, (C) Spine Research Institute of San Diego, 1999.

Figure 1.6. Lesions reported in CAD trauma. Segments illustrated in black portion of inset.
PICTURE 18